Recent multi-view diffusion (MVD) methods have utilized the generative capabilities of 2D image diffusion models to produce multi-view images from a single-view input. However, existing approaches often depend on dense crossview attention layers, which hinder scalability and fidelity due to their high computational costs. In this paper, we propose CTR3D, a novel method that incorporates token reduction in multi-view attention layers to efficiently generate dense, high-resolution multi-view images without restricting the camera viewpoints of the generated views. Our approach is designed into three key steps: redundancy removal, attention interaction, and token recovery. These steps leverage lightweight, projection-based techniques for multi-view token reduction and recovery, significantly improving the computational efficiency of MVD. By reducing the number of tokens in attention layers while preserving multi-view consistency, our model achieves state-of-the-art performance in novel view synthesis and 3D reconstruction while keeping efficiency for generation of dense high-resolution images and normals. Experimental results demonstrate that our method surpasses existing approaches, providing a more efficient and effective solution for multi-view generation. https://github.com/HKUST-SAIL/CTR3D

Orthodontic treatment needs regular tooth alignment checks, but current methods depend on clinic visits, limiting remote care. With the emergence of 3D Gaussian Splatting (3DGS), realistic novel views can be synthesized, making it possible for clinicians to remotely monitor orthodontic conditions. However, using only five intraoral images with unknown camera poses and dynamic lighting presents major challenges in dental applications. To address these challenges, we propose DentalGS, an enhanced 3DGS framework capable of synthesizing novel intraoral views from five post-orthodontic intraoral images and pre-orthodontic intraoral scan (IOS) data as prior, without camera poses. Our method initializes a Gaussian point cloud labeled with ISO-FDI tooth classes based on the patient’s pre-orthodontic IOS data, then estimates camera poses through iterative optimization. We introduce a Progressive Pair Generation Strategy as a data augmentation method that generates damage–repair image pairs to train a RepairNet, aiming to restore degraded geometry and appearance caused by the limited number of intraoral images. Additionally, we introduce a Lighting-Aware 3DGS inspired by physical reflectance properties to mitigate the effects of dynamic lighting conditions. Experimental results show that our method produces high-quality novel views while preserving geometric structure even under extreme viewpoints, offering an efficient and reliable solution for 3D tooth visualization in remote orthodontic monitoring.

3D Gaussian Splatting (3DGS) achieves impressive quality and rendering speed, but with millions of 3D Gaussians and significant storage and transmission costs. In this paper, we aim to develop a simple yet effective method called NeuralGS that compresses the original 3DGS into a compact representation. Our observation is that neural fields like NeRF can represent complex 3D scenes with Multi-Layer Perceptron (MLP) neural networks using only a few megabytes. Thus, NeuralGS effectively adopts the neural field representation to encode the attributes of 3D Gaussians with MLPs, only requiring a small storage size even for a large-scale scene. To achieve this, we adopt a clustering strategy and fit the Gaussians within each cluster using different tiny MLPs, based on importance scores of Gaussians as fitting weights. We experiment on multiple datasets, achieving a 91$\times$ average model size reduction without harming the visual quality.

Multi-view video reconstruction plays a vital role in computer vision, enabling applications in film production, virtual reality, and motion analysis. While recent advances such as 3D Gaussian Splatting have demonstrated impressive capabilities in dynamic scene reconstruction, they typically rely on the assumption that input video streams are temporally synchronized. However, in real-world scenarios, this assumption often fails due to factors like camera trigger delays, frame rate discrepancies, or independent recording setups, leading to temporal misalignment across views and reduced reconstruction quality. To address this challenge, a novel temporal alignment strategy is proposed for high-quality 4DGS reconstruction from unsynchronized multi-view videos. Our method features a coarse-to-fine alignment module that estimates and compensates for each camera's time shift. The method first determines a coarse, frame-level offset and then refines it to achieve sub-frame accuracy. This strategy can be integrated as a plug-and-play module into existing 4DGS frameworks, enhancing their robustness when handling asynchronous data. Experiments show that this approach effectively processes temporally misaligned videos and significantly enhances baseline methods.

Given a collection of points in $\mathbb {R}^{3}$R3, KD-Tree and R-Tree are well-known nearest neighbor search (NNS) algorithms that rely on spatial partitioning and indexing techniques. However, when the query point is far from the data points or the data points inherently represent a 2-manifold surface, their query performance may degrade. To address this, we propose a novel dynamic programming technique that precomputes a Directed Acyclic Graph (DAG) to encode the proximity structure between data points. More specifically, the DAG captures how the proximity structure evolves during the incremental construction of the Voronoi diagram of the data points. Experimental results demonstrate that our method achieves a speed increase of 1-10x. Furthermore, our algorithm demonstrates significant practical value in diverse applications. We validated its effectiveness through extensive testing in four key applications: Point-to-Mesh Distance Queries, Iterative Closest Point (ICP) Registration, Density Peak Clustering, and Point-to-Segments Distance Queries. A particularly notable feature of our approach is its unique ability to efficiently identify the nearest neighbor among the first $k$k points in the point cloud, a capability that enables substantial acceleration in low-dimensional applications like Density Peak Clustering. As a natural extension of our incremental construction process, our method can also be readily adapted for farthest-point sampling tasks. These experimental results across multiple domains underscore the broad applicability and practical importance of our approach.

Surface parameterization is a fundamental geometry processing task, laying the foundations for the visual presentation of 3D assets and numerous downstream shape analysis scenarios. Conventional parameterization approaches demand high-quality mesh triangulation and are restricted to certain simple topologies unless additional surface cutting and decomposition are provided. In practice, the optimal configurations (e.g., type of parameterization domains, distribution of cutting seams, number of mapping charts) may vary drastically with different surface structures and task characteristics, thus requiring more flexible and controllable processing pipelines. To this end, this paper introduces FlexPara, an unsupervised neural optimization framework to achieve both global and multi-chart surface parameterizations by establishing point-wise mappings between 3D surface points and adaptively-deformed 2D UV coordinates. We ingeniously design and combine a series of geometrically-interpretable sub-networks, with specific functionalities of cutting, deforming, unwrapping, and wrapping, to construct a bi-directional cycle mapping framework for global parameterization without the need for manually specified cutting seams. Furthermore, we construct a multi-chart parameterization framework with adaptively-learned chart assignment. Extensive experiments demonstrate the universality, superiority, and inspiring potential of our neural surface parameterization paradigm.

Accurately estimating human internal states, such as personality traits or behavioral patterns, is critical for enhancing the effectiveness of human–robot interaction, particularly in multi-agent settings. These insights are key in applications ranging from social navigation to autism diagnosis. However, prior methods are limited by scalability and passive observation, making real-time estimation in complex, multi-human settings difficult. In this work, we propose a practical method for active human personality estimation in crowds, with a focus on applications related to Autism Spectrum Disorder (ASD). Our method combines a personality-conditioned behavior model, based on the Eysenck 3-Factor theory, with an active robot information-gathering policy that triggers human behaviors through a receding-horizon planner. The robot’s belief about human personality is then updated via Bayesian inference. We demonstrate the effectiveness of our approach through proof-of-concept studies in simulation, user studies with typical adults, and preliminary experiments involving participants with ASD. Our results show that our method can scale to tens of humans and reduce personality estimation error by 29.2% and uncertainty by 79.9% in simulation compared to the passive baseline. User studies with typical adults confirm the method’s ability to generalize across complex personality distributions. Additionally, we explore its application in autism-related scenarios, demonstrating that the method can identify the difference between neurotypical and autistic behavior. The results suggest that our framework could serve as a foundation for future ASD-specific applications.

Orthodontic motion planning plays a crucial role in digital orthodontics by predicting tooth motion sequences to assist dentists in formulating treatment plans efficiently. Most prior work generates the entire intermediate tooth motion sequence given the initial and target tooth alignments. In practice, only the initial alignment of the patient is obtained. However, no existing method can predict the complete motion sequence using only the initial tooth alignment. To address this gap, we propose OrthoDiff, a novel target-free framework that uses only initial tooth alignment through a progressive generation strategy. This strategy generates tooth motion sequences by decomposing the entire motion sequence into multi-level motions, progressively constraining the inference space and reducing the complexity of target-free planning from coarse to fine. Moreover, we design a hierarchical diffusion transformer as the backbone of OrthoDiff, which treats tooth alignment as a sequence of tooth tokens and fully leverages the topological prior knowledge of the dental model. Through extensive evaluations, we demonstrate that our method significantly outperforms state-of-the-art techniques in target-free tooth motion generation. Ablation studies further confirm the efficacy of key components in our network design. Meanwhile, we also achieve state-of-the-art results in tooth target alignment prediction, benefiting from our framework. The code and data will be publicly available at https://github.com/Intelligent-Orthodontics/OrthoDiff.github.io.

We introduce NeuBase , a neural parametric surface representation that both accurately fits target surfaces with fine geometric detail and supports intuitive real time surface deformation. NeuBase consists of a Catmull-Clark subdivision base surface and an offset field defined by a set of neural basis functions encoded via a neural map. By construction, NeuBase surfaces exhibit four fundamental geometric properties, i.e., linearity, locality, smoothness, and affine equivariance, enabling real-time, direct manipulation without retraining the neural network. In addition, we propose a scalable neural map that maintains memory efficiency even for complex shapes with dense control meshes. Experiments on a large-scale dataset demonstrate that our method achieves better fitting accuracy than state-of-the-art neural parametric surface representations.

The Medial Axis Transform (MAT) is a complete shape descriptor capable of reconstructing the geometry of the original domain. A high-quality MAT should not only facilitate high-fidelity reconstruction but also capture structural features—for instance, by aligning the MAT boundary with the locus of rolling ball centers within fillet regions. However, computing such an ideal MAT remains a significant challenge, particularly when the input is a discrete triangle mesh. In this paper, we follow the established technical pipeline of initializing the MAT via a 3D Voronoi diagram of surface samples and subsequently simplifying the Voronoi structure through a QEM-like scheme. Our key insight is to explicitly track the correspondence between MAT vertices and surface regions throughout the progressive simplification process, ensuring that the resulting MAT triangles accurately reflect the intrinsic symmetries between surface patches. We translate these geometric requirements into a suite of priority control strategies that govern the sequencing of edge collapses. Through extensive evaluation against state-of-the-art MAT algorithms, we validate the strong performance of our approach regarding runtime efficiency, structural alignment, boundary regularity, triangle quality, and robustness to noise. Our resulting MATs remain highly expressive for both articulated shapes and CAD models, even under extreme simplification—effectively capturing the global structure of complex geometries with only a few hundred vertices. Finally, we showcase the utility of our approach through two potential applications: capturing the locus of rolling ball centers within fillet regions, a structural capability not previously demonstrated in the existing literature, and surface extraction from unsigned distance fields, where the medial axis of the є -isosurface naturally yields a clean single-layer result. Source code is available at https://github.com/sssomeone/structural-mat.

k -nearest neighbor ( k -NN) search is a fundamental primitive in geometry processing and computer graphics. While spatial partitioning structures such as kd -trees are standard, they are often manifold-blind, failing to exploit the intrinsic low-dimensional structure of points sampled from 2-manifolds. Recent advances in dynamic programming-based nearest neighbor search (DP-NNS) leverage incrementally constructed Voronoi diagrams to accelerate queries, where each site p maintains a list of successors that progressively refine its Voronoi cell. However, DP-NNS is restricted to single nearest neighbor ( k = 1) searches, precluding their adoption in applications that require local neighborhood statistics. In this paper, we generalize the DP-NNS framework to support arbitrary k -NN queries for manifold-aligned data. Our approach is founded on the geometric observation that if p i is the nearest neighbor of a query q in P , then the second nearest neighbor of q must reside either within the prefix set P 1: i -1 = [ p 1 , ..., p i-1 } or within p i 's successor list. By recursively extending this principle, we introduce Manifold k -NN, a recursive algorithmic scheme that significantly outperforms conventional kd -trees for manifold-aligned data. Our method achieves a 1×-10× speedup in volume-to-surface query scenarios and inherently supports dynamic prefix queries—enabling k -NN searches within any subset P 1: m ( m ≤ n ) with zero overhead. Furthermore, we extend the framework to support point deletion via local Delaunay updates, providing a complete suite of dynamic operations for point set modification. Comprehensive experiments on diverse geometric datasets demonstrate the efficiency and broad applicability of our approach for modern graphics pipelines. Source code is available at https://github.com/sssomeone/manifold-knn.

Cages are fundamental structures in computer graphics, serving as versatile proxies for a wide range of applications. A high-quality cage must balance two competing objectives: minimizing the face count to ensure simplicity, and maximizing tightness to maintain high geometric fidelity to the input mesh. In this paper, we propose PR-Cage, a nested optimization framework for automated cage generation. For the outer control layer, we introduce a thickness parameter τ that defines a feasibility region; the evolving cage is guided by the τ -offset surface. We observe that an optimal balance between simplicity and tightness is achievable by progressively relaxing the parameter τ via a staircase schedule. For the inner iterations, we extend the traditional Quadric Error Metric (QEM) framework by incorporating rigorous linear inequality constraints to suppress triangle degeneration and prevent normal flips. Our algorithm relies exclusively on the atomic operations of edge collapses and edge flips, resulting in high computational efficiency and robustness. Comparative experiments on public datasets demonstrate that PR-Cage consistently outperforms existing methods, achieving extreme simplification while maintaining high adherence to the underlying geometry; see the teaser figure. Due to these favorable properties, we demonstrate the utility of our method in several downstream applications, such as contact simulation and deformation, where PR-Cage exhibits significant advantages in both quality and performance.

Recent advancements in autoregressive transformers have demonstrated remarkable potential for generating artist-quality meshes. However, the token ordering strategies employed by existing methods typically fail to meet professional artist standards, where coordinate-based sorting yields inefficiently long sequences, and patch-based heuristics disrupt the continuous edge flow and structural regularity essential for high-quality modeling. To address these limitations, we propose Strips as Tokens ( SATO ), a novel framework with a token ordering strategy inspired by triangle strips. By constructing the sequence as a connected chain of faces that explicitly encodes UV boundaries, our method naturally preserves the organized edge flow and semantic layout characteristic of artist-created meshes. A key advantage of this formulation is its unified representation, enabling the same token sequence to be decoded into either a triangle or quadrilateral mesh. This flexibility facilitates joint training on both data types: large-scale triangle data provides fundamental structural priors, while high-quality quad data enhances the geometric regularity of the outputs. Extensive experiments demonstrate that SATO consistently outperforms prior methods in terms of geometric quality, structural coherence, and UV segmentation.

In this paper, we study an under-explored but important factor of diffusion generative models, i.e., the combinatorial complexity. Data samples are generally high-dimensional, and for various structured generation tasks, additional attributes are combined to associate with data samples. We show that the space spanned by the combination of dimensions and attributes can be insufficiently covered by existing training schemes of diffusion generative models, potentially limiting test time performance. We present a simple fix to this problem by constructing stochastic processes that fully exploit the combinatorial structures, hence the name ComboStoc. Using this simple strategy, we show that network training is significantly accelerated across diverse data modalities, including images and 3D structured shapes. Moreover, ComboStoc enables a new way of test time generation which uses asynchronous time steps for different dimensions and attributes, thus allowing for varying degrees of control over them. Our code is available at: https://github.com/Xrvitd/ComboStoc.

Implicit neural representations are powerful for geometric modeling, but their practical use is often limited by the high computational cost of network evaluations. We observe that implicit representations require progressively lower accuracy as query points move farther from the target surface, and that even within the same iso-surface, representation difficulty varies spatially with local geometric complexity. However, conventional neural implicit models evaluate all query points with the same network depth and computational cost, ignoring this spatial variation and thereby incurring substantial computational waste. Motivated by this observation, we propose an efficient neural implicit geometry representation framework with spatially adaptive network depth (SAND). SAND leverages a volumetric network-depth map together with a tailed multi-layer perceptron (T-MLP) to model implicit representation. The volumetric depth map records, for each spatial region, the network depth required to achieve sufficient accuracy, while the T-MLP is a modified MLP designed to learn implicit functions such as signed distance functions, where an output branch, referred to as a tail, is attached to each hidden layer. This design allows network evaluation to terminate adaptively without traversing the full network and directs computational resources to geometrically important and complex regions, improving efficiency while preserving high-fidelity representations. Extensive experimental results demonstrate that our approach can significantly improve the inference-time query speed of implicit neural representations.

Piecewise parametric surfaces have long been established as prevalent geometric representations; however, they often require surface refinement or sophisticated quadrangulation to accurately represent complex geometries. Geometric deep learning has shown that neural networks can provide greater representational power than conventional methods. Nevertheless, approaches using a single parametric surface for shape fitting struggle to capture fine-grained geometric details, while multi-patch methods fail to ensure seamless connections between adjacent patches. We present Neural Piecewise Parametric Surfaces ( NeuPPS ), the first piecewise neural surface representation that allows for coarse patch layouts composed of arbitrary n -sided surface patches to model complex surface geometries with high precision, offering enhanced flexibility compared with traditional parametric surfaces. This new surface representation guarantees, by construction, the continuity between adjacent patches, a property that other neural patch-based approaches cannot ensure. Two novel components are introduced: a learnable feature complex and a continuous mapping function approximated by multi-layer perceptrons (MLPs). We apply the proposed NeuPPS to surface fitting and shape space learning tasks. Extensive experiments demonstrate the advantages of NeuPPS over traditional parametric representations and existing patch-based learning approaches.

Generating high-quality character animation videos is a fascinating yet challenging task. Existing methods use geometry guidance signals like skeletons, normal maps, or depth maps in a diffusion model to generate character videos from a single reference image. Although these approaches have shown encouraging results, they solely rely on cross attention layers to extract geometry guidance which inevitably leads to temporal inconsistencies and reduced quality. In this paper, we present a novel framework AniFeats to generate high-quality character animation videos. In contrast to existing methods, our key insight is to incorporate explicit features on 3D character meshes during the video generation to achieve significantly improved temporal consistency. Specifically, AniFeats extracts detailed features from the reference image, projects them onto 3D feature meshes based on SMPL-X, and utilizes rendered feature maps from the animated 3D feature meshes as guidance throughout the generation process. This approach directly links local patterns in the input image to those in the output video, effectively strengthening temporal coherence. Extensive experiments demonstrate that AniFeats generates high-quality, temporally consistent character animations with remarkably enhanced realism.

Extracting high-fidelity mesh surfaces from Signed Distance Fields (SDFs) has become a fundamental operation in geometry processing. Despite significant progress over the past decades, key challenges remain-namely, how to automatically capture the intricate geometric and topological structures encoded in the zero level set of SDFs. In this paper, we present a novel isosurface extraction algorithm that introduces two key innovations: 1) An incrementally constructed power diagram through the addition of sample points, which enables repeated updates to the extracted surface via its dual-regular Delaunay tetrahedralization; and 2) An adaptive point insertion strategy that identifies regions exhibiting the greatest discrepancy between the current mesh and the underlying continuous surface. As Fig. 1 shows, our framework progressively refines the extracted mesh with minimal computational cost until it sufficiently approximates the underlying surface. Experimental results demonstrate that our approach outperforms state-of-the-art methods, particularly for models with intricate geometric variations and complex topologies.

Gaussian Splatting demonstrates impressive results in multi-view reconstruction based on Gaussian explicit representations. However, the current Gaussian primitives only have a single view-dependent color and an opacity to represent the appearance and geometry of the scene, resulting in a non-compact representation. In this paper, we introduce a new method called SVGS (Spatially Varying Gaussian Splatting) that utilizes spatially varying colors and opacity in a single Gaussian primitive to improve its representation ability. We have implemented bilinear interpolation, movable kernels, and tiny neural networks as spatially varying functions. SVGS employs 2D Gaussian surfels as primitives, which significantly enhances novel-view synthesis while maintaining high-quality geometric reconstruction. This approach is particularly effective in practical applications, as scenes combining complex textures with relatively simple geometry occur frequently in real-world environments. Quantitative and qualitative experimental results demonstrate that all three functions outperform the baseline, with the best movable kernels achieving superior novel view synthesis performance on multiple datasets, highlighting the strong potential of spatially varying functions.

Conventional production workflow of high-precision mesh assets necessitates a cumbersome and laborious process of manual sculpting by specialized 3D artists/modelers. The recent years have witnessed remarkable advances in AI-empowered 3D content creation for generating plausible structures and intricate appearances from images or text prompts. However, synthesizing realistic surface details still poses great challenges, and enhancing the geometry fidelity of existing lower-quality 3D meshes (instead of image/text-to-3D generation) remains an open problem. In this paper, we introduce SuperCarver, a 3D geometry super-resolution pipeline for supplementing texture-consistent surface details onto a given coarse mesh. We start by rendering the original textured mesh into the image domain from multiple viewpoints. To achieve detail boosting, we construct a deterministic prior-guided normal diffusion model, which is fine-tuned on a carefully curated dataset of paired detail-lacking and detail-rich normal map renderings. To update mesh surfaces from potentially imperfect normal map predictions, we design a noise-resistant inverse rendering scheme through deformable distance field. Experiments demonstrate that our SuperCarver is capable of generating realistic and expressive surface details depicted by the actual texture appearance, making it a powerful tool to both upgrade historical low-quality 3D assets and reduce the workload of sculpting high-poly meshes.

Offset surfaces, defined as the Minkowski sum of a base surface and a rolling ball, play a crucial role in geometry processing, with applications ranging from coverage motion planning to brush modeling. While considerable progress has been made in computing constant-radius offset surfaces, computing variable-radius offset surfaces remains a challenging problem. In this paper, we present OffsetCrust, a novel framework that efficiently addresses the variable-radius offsetting problem by computing a power diagram. Let ${\mathcal {R}}$R denote the radius function defined on the base surface $\mathcal {S}$S. The power diagram is constructed from contributing sites, consisting of carefully sampled base points on $\mathcal {S}$S and their corresponding off-surface points, displaced along ${\mathcal {R}}$R-dependent directions. In the constant-radius case only, these displacement directions align exactly with the surface normals of $\mathcal {S}$S. Moreover, our method mitigates the misalignment issues commonly seen in crust-based approaches through a lightweight fine-tuning procedure. We validate the accuracy and efficiency of OffsetCrust through extensive experiments, and demonstrate its practical utility in applications such as reconstructing original boundary surfaces from medial axis transform (MAT) representations.

Diffusion-based 3D generation has made remarkable progress in recent years. However, existing 3D generative models often produce overly dense and unstructured meshes, which stand in stark contrast to the compact, structured, and sharply-edged Computer-Aided Design (CAD) models crafted by human designers. To address this gap, we introduce CADDreamer, a novel approach for generating boundary representations (B-rep) of CAD objects from a single image. CADDreamer employs a primitive-aware multi-view diffusion model that captures both local geometric details and high-level structural semantics during the generation process. By encoding primitive semantics into the color domain, the method leverages the strong priors of pre-trained diffusion models to align with well-defined primitives. This enables the inference of multi-view normal maps and semantic maps from a single image, facilitating the reconstruction of a mesh with primitive labels. Furthermore, we introduce geometric optimization techniques and topology-preserving extraction methods to mitigate noise and distortion in the generated primitives. These enhancements result in a complete and seamless B-rep of the CAD model. Experimental results demonstrate that our method effectively recovers high-quality CAD objects from single-view images. Compared to existing 3D generation techniques, the B-rep models produced by CADDreamer are compact in representation, clear in structure, sharp in edges, and watertight in topology.

Our work aims to reconstruct hand-object interactions from a single-view image, which is a fundamental but ill-posed task. Unlike methods that reconstruct from videos, multiview images, or predefined 3D templates, single-view reconstruction faces significant challenges due to inherent ambiguities and occlusions. These challenges are further amplified by the diverse nature of hand poses and the vast variety of object shapes and sizes. Our key insight is that current foundational models for segmentation, inpainting, and 3D reconstruction robustly generalize to in-the-wild images, which could provide strong visual and geo-metric priors for reconstructing hand-object interactions. Specifically, given a single image, we first design a novel pipeline to estimate the underlying hand pose and object shape using off-the-shelf large models. Furthermore, with the initial reconstruction, we employ a prior-guided optimization scheme, which optimizes hand pose to comply with 3D physical constraints and the 2D input image content. We perform experiments across several datasets and show that our method consistently outperforms baselines and faithfully reconstructs a diverse set of hand-object interactions. Here is the link of our project page: https://lym29.github.io/EasyHOI-page/.

Recent developments in monocular depth estimation methods enable high-quality depth estimation of single-view images but fail to estimate consistent video depth across different frames. Very recent works address this problem by applying a video diffusion model to generate video depth conditioned on the input video, which is training-expensive and can only produce scale-invariant depth values without camera poses. In this paper, we propose a novel video-depth estimation method called Align3R to estimate temporally consistent depth maps for a dynamic video. Our key idea is to utilize the recent DUSt3R model to align estimated monocular depth maps of different timesteps. First, we fine-tune the DUSt3R model with additional estimated monocular depth as inputs for the dynamic scenes. Then, we apply optimization to reconstruct both depth maps and camera poses. Extensive experiments demonstrate that Align3R estimates consistent video depth and camera poses for a monocular video with superior performance than baseline methods.

Deep neural networks (DNNs) often exhibit out-of-distribution (OOD) overconfidence, producing overly confident predictions on OOD samples. We attribute this issue to the inherent over-complexity of DNNs and investigate two key aspects: capacity and nonlinearity. First, we demonstrate that reducing model capacity through knowledge distillation can effectively mitigate OOD overconfidence. Second, we show that selectively reducing nonlinearity by removing ReLU operations further alleviates the issue. Building on these findings, we present a practical guide to model simplification, combining both strategies to significantly reduce OOD overconfidence. Extensive experiments validate the effectiveness of this approach in mitigating OOD overconfidence and demonstrate its superiority over state-of-the-art methods. Additionally, our simplification strategies can be combined with existing OOD detection techniques to further enhance OOD detection performance.

3D Gaussian Splatting (3DGS) is a prevailing technique to reconstruct large-scale 3D scenes from multiview images for novel view synthesis, like a room, a block, and even a city. Such large-scale scenes are not static with changes constantly happening in these scenes, like a new building being built or a new decoration being set up. To keep the reconstructed 3D Gaussian fields up-to-date, a naive way is to reconstruct the whole scene after changing, which is extremely costly and inefficient. In this paper, we propose a new method called GauUpdate that allows partially updating an old 3D Gaussian field with new objects from a new 3D Gaussian field. However, simply inserting the new objects leads to inconsistent appearances because the old and new Gaussian fields may have different lighting environments from each other. GauUpdate addresses this problem by applying inverse rendering techniques in the 3DGS to recover both the materials and environmental lights. Based on the materials and lighting, we relight the new objects in the old 3D Gaussian field for consistent global illumination. For an accurate estimation of the materials and lighting, we put additional constraints on the materials and lighting conditions, that these two fields share the same materials but different environment lights, to improve their qualities. We conduct experiments on both synthetic scenes and real-world scenes to evaluate GauUpdate, which demonstrate that GauUpdate achieves realistic object insertion in 3D Gaussian fields with consistent appearances.

In this paper, we propose VistaDream a novel framework to reconstruct a 3D scene from a single-view image. Recent diffusion models enable generating high-quality novel-view images from a single-view input image. Most existing methods only concentrate on building the consistency between the input image and the generated images while losing the consistency between the generated images. VistaDream addresses this problem by a two-stage pipeline. In the first stage, VistaDream begins with building a global coarse 3D scaffold by zooming out a little step with inpainted boundaries and an estimated depth map. Then, on this global scaffold, we use iterative diffusion-based RGB-D inpainting to generate novel-view images to inpaint the holes of the scaffold. In the second stage, we further enhance the consistency between the generated novel-view images by a novel training-free Multiview Consistency Sampling (MCS) that introduces multi-view consistency constraints in the reverse sampling process of diffusion models. Experimental results demonstrate that without training or fine-tuning existing diffusion models, VistaDream achieves consistent and high-quality novel view synthesis using just single-view images and outperforms baseline methods by a large margin. The code, videos, and interactive demos are available at https://vistadream-project-page.github.io/.

In this paper, we present a 3D visual grounding method called CityAnchor for localizing an urban object in a city-scale point cloud. Recent developments in multiview reconstruction enable us to reconstruct city-scale point clouds but how to conduct visual grounding on such a large-scale urban point cloud remains an open problem. Previous 3D visual grounding system mainly concentrates on localizing an object in an image or a small-scale point cloud, which is not accurate and efficient enough to scale up to a city-scale point cloud. We address this problem with a multi-modality LLM which consists of two stages, a coarse localization and a fine-grained matching. Given the text descriptions, the coarse localization stage locates possible regions on a projected 2D map of the point cloud while the fine-grained matching stage accurately determines the most matched object in these possible regions. We conduct experiments on the CityRefer dataset and a new synthetic dataset annotated by us, both of which demonstrate our method can produce accurate 3D visual grounding on a city-scale 3D point cloud.

In this paper, we propose MoDGS, a new pipeline to render novel-view images in dynamic scenes using only casually captured monocular videos. Previous monocular dynamic NeRF or Gaussian Splatting methods strongly rely on the rapid movement of input cameras to construct multiview consistency but fail to reconstruct dynamic scenes on casually captured input videos whose cameras are static or move slowly. To address this challenging task, MoDGS adopts recent single-view depth estimation methods to guide the learning of the dynamic scene. Then, a novel 3D-aware initialization method is proposed to learn a reasonable deformation field and a new robust depth loss is proposed to guide the learning of dynamic scene geometry. Comprehensive experiments demonstrate that MoDGS is able to render high-quality novel view images of dynamic scenes from just a casually captured monocular video, which outperforms baseline methods by a significant margin. Project page: https://MoDGS.github.io

Reconstructing 3D hand-face interactions with deformations from a single image is a challenging yet crucial task with broad applications in AR, VR, and gaming. The challenges stem from self-occlusions during single-view hand-face interactions, diverse spatial relationships between hands and face, complex deformations, and the ambiguity of the single-view setting. The previous state-of-the-art, Decaf, employs a global fitting optimization guided by contact and deformation estimation networks trained on studio-collected data with 3D annotations. However, Decaf suffers from a time-consuming optimization process and limited generalization capability due to its reliance on 3D annotations of hand-face interaction data. To address these issues, we present DICE, the first end-to-end method for Deformation-aware hand-face Interaction reCovEry from a single image. DICE estimates the poses of hands and faces, contacts, and deformations simultaneously using a Transformer-based architecture. It features disentangling the regression of local deformation fields and global mesh vertex locations into two network branches, enhancing deformation and contact estimation for precise and robust hand-face mesh recovery. To improve generalizability, we propose a weakly-supervised training approach that augments the training set using in-the-wild images without 3D ground-truth annotations, employing the depths of 2D keypoints estimated by off-the-shelf models and adversarial priors of poses for supervision. Our experiments demonstrate that DICE achieves state-of-the-art performance on a standard benchmark and in-the- wild data in terms of accuracy and physical plausibility. Additionally, our method operates at an interactive rate (20 fps) on an Nvidia 4090 GPU, whereas Decaf requires more than 15 seconds for a single image. The code will be available at: https://github.com/Qingxuan-Wu/DICE.

In recent years, simultaneous learning of multiple dense prediction tasks with partially annotated label data has emerged as an important research area. Previous works primarily focus on leveraging cross-task relations or conducting adversarial training for extra regularization, which achieve promising performance improvements, while still suffering from the lack of direct pixel-wise supervision and extra training of heavy mapping networks. To effectively tackle this challenge, we propose a novel approach to optimize a set of compact learnable hierarchical task tokens, including global and fine-grained ones, to discover consistent pixel-wise supervision signals in both feature and prediction levels. Specifically, the global task tokens are designed for effective cross-task feature interactions in a global context. Then, a group of fine-grained task-specific spatial tokens for each task is learned from the corresponding global task tokens. It is embedded to have dense interactions with each task-specific feature map. The learned global and local fine-grained task tokens are further used to discover pseudo task-specific dense labels at different levels of granularity, and they can be utilized to directly supervise the learning of the multi-task dense prediction framework. Extensive experimental results on challenging NYUD-v2, Cityscapes, and PASCAL Context datasets demonstrate significant improvements over existing state-of-the-art methods for partially annotated multi-task dense prediction.

The integration of language and 3D perception is critical for embodied AI and robotic systems to perceive, understand, and interact with the physical world. Spatial reasoning, a key capability for understanding spatial relationships between objects, remains underexplored in current 3D vision-language research. Existing datasets often mix semantic cues (e.g., object name) with spatial context, leading models to rely on superficial shortcuts rather than genuinely interpreting spatial relationships. To address this gap, we introduce Surprise3D, a novel dataset designed to evaluate language-guided spatial reasoning segmentation in complex 3D scenes. Surprise3D consists of more than 200k vision language pairs across 900+ detailed indoor scenes from ScanNet++ v2, including more than 2.8k unique object classes. The dataset contains 89k+ human-annotated spatial queries deliberately crafted without object name, thereby mitigating shortcut biases in spatial understanding. These queries comprehensively cover various spatial reasoning skills, such as relative position, narrative perspective, parametric perspective, and absolute distance reasoning. Initial benchmarks demonstrate significant challenges for current state-of-the-art expert 3D visual grounding methods and 3D-LLMs, underscoring the necessity of our dataset and the accompanying 3D Spatial Reasoning Segmentation (3D-SRS) benchmark suite. Surprise3D and 3D-SRS aim to facilitate advancements in spatially aware AI, paving the way for effective embodied interaction and robotic planning.

Enabling virtual humans to dynamically and realistically respond to diverse auditory stimuli remains a key challenge in character animation, demanding the integration of perceptual modeling and motion synthesis. Despite its significance, this task remains largely unexplored. Most previous works have primarily focused on mapping modalities like speech, audio, and music to generate human motion. As of yet, these models typically overlook the impact of spatial features encoded in spatial audio signals on human motion. To bridge this gap and enable high-quality modeling of human movements in response to spatial audio, we introduce the first comprehensive "Spatial Audio-Driven Human Motion" (SAM) dataset, which contains diverse and high-quality spatial audio and motion data. For benchmarking, we develop a simple yet effective diffusion-based generative framework for human "MOtion generation driven by SPatial Audio," termed MOSPA, which faithfully captures the relationship between body motion and spatial audio through an effective fusion mechanism. Once trained, MOSPA can generate diverse realistic human motions conditioned on varying spatial audio inputs. We perform a thorough investigation of the proposed dataset and conduct extensive experiments for benchmarking, where our method achieves state-of-the-art performance on this task. Our code and model are publicly available at https://github.com/xsy27/Mospa-Acoustic-driven-Motion-Generation.git

We propose a novel optimization framework for computing the medial axis transform that simultaneously preserves the medial structure and ensures high medial mesh quality. The medial structure, consisting of interconnected sheets, seams, and junctions, provides a natural volumetric decomposition of a 3D shape. Our method introduces a structure-aware, particle-based optimization pipeline guided by the restricted power diagram (RPD), which partitions the input volume into convex cells whose dual encodes the connectivity of the medial mesh. Structure-awareness is enforced through a spherical quadratic error metric (SQEM) projection that constrains the movement of medial spheres, while a Gaussian kernel energy encourages an even spatial distribution. Compared to feature-preserving methods such as MATFP [Wang et al. 2022] and MATTopo [Wang et al. 2024b], our approach produces cleaner medial structures with significantly improved mesh quality. In contrast to voxel-based, point-cloud-based, and variational methods, our framework is the first to integrate structural awareness into the optimization process, yielding medial meshes with explicit structural decomposition, topological correctness, and geometric fidelity. Our code is available at our project website.

Existing single-view 3D generative models typically adopt multiview diffusion priors to reconstruct object surfaces, yet they remain prone to inter-view inconsistencies and are unable to faithfully represent complex internal structure or nontrivial topologies. In particular, we encode geometry information by projecting it onto a bounding sphere and unwrapping it into a compact and structural multi-layer 2D Spherical Projection (SP) representation. Operating solely in the image domain, SPGen offers three key advantages simultaneously: (1) Consistency. The injective SP mapping encodes surface geometry with a single viewpoint which naturally eliminates view inconsistency and ambiguity; (2) Flexibility. Multi-layer SP maps represent nested internal structures and support direct lifting to watertight or open 3D surfaces; (3) Efficiency. The image-domain formulation allows the direct inheritance of powerful 2D diffusion priors and enables efficient finetuning with limited computational resources. Extensive experiments demonstrate that SPGen significantly outperforms existing baselines in geometric quality and computational efficiency.

Abstract Swept volume computation—the determination of regions occupied by moving objects—is essential in graphics, robotics, and manufacturing. Existing approaches either explicitly track surfaces, suffering from robustness issues under complex interactions, or employ implicit representations that trade off geometric fidelity and face optimization difficulties. We propose a novel inversion of motion perspective: rather than tracking object motion, we fix the object and trace spatial points backward in time, reducing complex trajectories to efficiently linearizable point motions. Based on this, we introduce a multi‐field tetrahedral framework that maintains multiple distance fileds per element, preserving fine geometric details at trajectory intersections where single‐field methods fail. Our method robustly computes swept volumes for diverse motions, including translations and screw motions, and enables practical applications in path planning and collision detection.

This paper presents a novel approach for camera pose refinement based on neural radiance fields (NeRF) by introducing semantic feature consistency to enhance robustness. NeRF has been successfully applied to camera pose estimation by inverting the rendering process given an observed RGB image and an initial pose estimate. However, previous methods only adopted photometric consistency for pose optimization, which is prone to be trapped in local minima. To address this problem, we introduce semantic feature consistency into the existing framework. Specifically, we utilize high-level features extracted from a convolutional neural network (CNN) pre-trained for image recognition, and maintain consistency of such features between observed and rendered images during the optimization procedure. Unlike the color values at each pixel, these features contain rich semantic information shared within local regions and can be more robust to appearance changes from different viewpoints. Since it is computationally expensive to render a full image with NeRF for feature extraction from CNN, we propose an efficient way to estimate the features of individually rendered pixels by projecting them to a nearby reference image and interpolating its feature maps. Extensive experiments show that our method greatly outperforms the baseline method on both synthetic objects and real-world large indoor scenes, increasing the accuracy of pose estimation by over 6.4%.

Qualifying the discrepancy between 3D geometric models, which could be represented with either point clouds or triangle meshes, is a pivotal issue with board applications. Existing methods mainly focus on directly establishing the correspondence between two models and then aggregating point-wise distance between corresponding points, resulting in them being either inefficient or ineffective. In this paper, we propose DDM, an efficient, effective, robust, and differentiable distance metric for 3D geometry data. Specifically, we construct DDM based on the proposed implicit representation of 3D models, namely directional distance field (DDF), which defines the directional distances of 3D points to a model to capture its local surface geometry. We then transfer the discrepancy between two 3D geometric models as the discrepancy between their DDFs defined on an identical domain, naturally establishing model correspondence. To demonstrate the advantage of our DDM, we explore various distance metric-driven 3D geometric modeling tasks, including template surface fitting, rigid registration, non-rigid registration, scene flow estimation and human pose optimization. Extensive experiments show that our DDM achieves significantly higher accuracy under all tasks. As a generic distance metric, DDM has the potential to advance the field of 3D geometric modeling.

As commonly used implicit geometry representations, the signed distance function (SDF) is limited to modeling watertight shapes, while the unsigned distance function (UDF) is capable of representing various surfaces. However, its inherent theoretical shortcoming, i.e., the non-differentiability at the zero-level set, would result in sub-optimal reconstruction quality. In this paper, we propose the scaled-squared distance function (S2DF), a novel implicit surface representation for modeling arbitrary surface types. S2DF does not distinguish between inside and outside regions while effectively addressing the non-differentiability issue of UDF at the zero-level set. We demonstrate that S2DF satisfies a second-order partial differential equation of Monge-Ampere-type, allowing us to develop a learning pipeline that leverages a novel MongeAmpere regularization to directly learn S2DF from raw unoriented point clouds without supervision from ground-truth S2DF values. Extensive experiments across multiple datasets show that our method significantly outperforms state-of-the-art supervised approaches that require ground-truth surface information as supervision for training. The code will be publicly available at https://github.com/chuanxiang-yang/S2DF.

Medical image segmentation has seen great progress in recent years, largely due to the development of deep neural networks. However, unlike in computer vision, high-quality clinical data is relatively scarce, and the annotation process is often a burden for clinicians. As a result, the scarcity of medical data limits the performance of existing medical image segmentation models. In this paper, we propose a novel framework that integrates eye tracking information from experienced radiologists during the screening process to improve the performance of deep neural networks with limited data. Our approach, a grouped hierarchical network, guides the network to learn from its faults by using gaze information as weak supervision. We demonstrate the effectiveness of our framework on mammogram images, particularly for handling segmentation classes with large scale differences. We evaluate the impact of gaze information on medical image segmentation tasks and show that our method achieves better segmentation performance compared to state-of-the-art models. A robustness study is conducted to investigate the influence of distraction or inaccuracies in gaze collection. We also develop a convenient system for collecting gaze data without interrupting the normal clinical workflow. Our work offers novel insights into the potential benefits of integrating gaze information into medical image segmentation tasks.

Convolutional neural networks, in which each layer receives features from the previous layer(s) and then aggregates/abstracts higher level features from them, are widely adopted for image classification. To avoid information loss during feature aggregation/abstraction and fully utilize lower layer features, we propose a novel decision fusion module (DFM) for making an intermediate decision based on the features in the current layer and then fuse its results with the original features before passing them to the next layers. This decision is devised to determine an auxiliary category corresponding to the category at a higher hierarchical level, which can, thus, serve as category-coherent guidance for later layers. Therefore, by stacking a collection of DFMs into a classification network, the generated decision fusion network is explicitly formulated to progressively aggregate/abstract more discriminative features guided by these decisions and then refine the decisions based on the newly generated features in a layer-by-layer manner. Comprehensive results on four benchmarks validate that the proposed DFM can bring significant improvements for various common classification networks at a minimal additional computational cost and are superior to the state-of-the-art decision fusion-based methods. In addition, we demonstrate the generalization ability of the DFM to object detection and semantic segmentation.

Quadrilateral mesh generation plays a crucial role in numerical simulations within Computer-Aided Design and Engineering (CAD/E). Producing high-quality quadrangulation typically requires satisfying four key criteria. First, the quadrilateral mesh should closely align with principal curvature directions. Second, singular points should be strategically placed and effectively minimized. Third, the mesh should accurately conform to sharp feature edges. Lastly, quadrangulation results should exhibit robustness against noise and minor geometric variations. Existing methods generally involve first computing a regular cross field to represent quad element orientations across the surface, followed by extracting a quadrilateral mesh aligned closely with this cross field. A primary challenge with this approach is balancing the smoothness of the cross field with its alignment to pre-computed principal curvature directions, which are sensitive to small surface perturbations and often ill-defined in spherical or planar regions. To tackle this challenge, we propose NeurCross , a novel framework that simultaneously optimizes a cross field and a neural signed distance function (SDF), whose zero-level set serves as a proxy of the input shape. Our joint optimization is guided by three factors: faithful approximation of the optimized SDF surface to the input surface, alignment between the cross field and the principal curvature field derived from the SDF surface, and smoothness of the cross field. Acting as an intermediary, the neural SDF contributes in two essential ways. First, it provides an alternative, optimizable base surface exhibiting more regular principal curvature directions for guiding the cross field. Second, we leverage the Hessian matrix of the neural SDF to implicitly enforce cross field alignment with principal curvature directions, thus eliminating the need for explicit curvature extraction. Extensive experiments demonstrate that NeurCross outperforms the state-of-the-art methods in terms of singular point placement, robustness against surface noise and surface undulations, and alignment with principal curvature directions and sharp feature curves.

Cross fields play a critical role in various geometry processing tasks, especially for quad mesh generation. Existing methods for cross field generation often struggle to balance computational efficiency with generation quality, using slow per-shape optimization. We introduce CrossGen , a novel framework that supports both feed-forward prediction and latent generative modeling of cross fields for quad meshing by unifying geometry and cross field representations within a joint latent space. Our method enables extremely fast computation of high-quality cross fields of general input shapes, typically within one second without per-shape optimization. Our method assumes a point-sampled surface, also called a point-cloud surface , as input, so we can accommodate various surface representations by a straightforward point sampling process. Using an auto-encoder network architecture, we encode input point-cloud surfaces into a sparse voxel grid with fine-grained latent spaces, which are decoded into both SDF-based surface geometry and cross fields (see the teaser figure). We also contribute a dataset of models with both high-quality signed distance fields (SDFs) representations and their corresponding cross fields, and use it to train our network. Once trained, the network is capable of computing a cross field of an input surface in a feed-forward manner, ensuring high geometric fidelity, noise resilience, and rapid inference. Furthermore, leveraging the same unified latent representation, we incorporate a diffusion model for computing cross fields of new shapes generated from partial input, such as sketches. To demonstrate its practical applications, we validate CrossGen on the quad mesh generation task for a large variety of surface shapes. Experimental results demonstrate that CrossGen generalizes well across diverse shapes and consistently yields high-fidelity cross fields, thus facilitating the generation of high-quality quad meshes.

Hand-drawn vector sketches often contain implied lines, imprecise intersections, and unintended gaps, making it challenging to identify closed regions for colorization. These challenges become more pronounced as the number of strokes increases. In this paper, we present KISSColor, a novel method for inferring users' intended closed regions. Specifically, we propose intuitive stroke stretching by extending open strokes along tangent isolines of winding-number fields, which provably form geometrically aligned closed regions. Extending all open strokes can lead to overly fragmented regions due to redundant intersections. While a Mixed Integer Programming (MIP) formulation helps reduce redundancy, it is computationally expensive. To improve efficiency, we introduce kinetic stroke stretching, which grows all strokes simultaneously and prioritizes early intersections using a kinetic data structure. This approach preserves stylistic ambiguity for lines requiring long extensions. Based on the growth results, redundant regions are suppressed to minimize fragmentation. We conduct extensive experiments demonstrating the effectiveness of KISSColor, which generates more intuitive partitions, especially for imprecise sketches (see teaser figure). Our code and data will be released upon publication.

Humans possess the ability to master a wide range of motor skills, enabling them to quickly and flexibly adapt to the surrounding environment. Despite recent progress in replicating such versatile human motor skills, existing research often oversimplifies or inadequately captures the complex interplay between human body movements and highly dynamic environments, such as interactions with fluids. In this paper, we present a world model for Character-Fluid Coupling (CFC) for simulating human-fluid interactions via two-way coupling. We introduce a two-level world model which consists of a Physics-Informed Neural Network (PINN)-based model for fluid dynamics and a character world model capturing body dynamics under various external forces. This two-level world model adeptly predicts the dynamics of fluid and its influence on rigid bodies via force prediction, sidestepping the computational burden of fluid simulation and providing policy gradients for efficient policy training. Once trained, our system can control characters to complete high-level tasks while adaptively responding to environmental changes. We also present that the fluid initiates emergent behaviors of the characters, enhancing motion diversity and interactivity. Extensive experiments underscore the effectiveness of CFC, demonstrating its ability to produce high-quality, realistic human-fluid interaction animations.

Filleting is a fundamental operation in CAD systems, akin to a ball rolling between two adjacent surface patches, resulting in a seamless connection. The reverse process, which we refer to as DeFillet in this paper, is crucial for CAE analysis and secondary design phases. However, it presents significant challenges, particularly when the input data originates from surface reconstruction or discretization processes. Our DeFillet algorithm is inspired by the observation that the rolling-ball center defines an osculating sphere, while the Voronoi diagram of surface samples provides sufficiently many rolling-ball center candidates. By leveraging this insight, we compute a transformation between the Voronoi vertices and the surface samples, enabling the efficient identification of fillet regions. Subsequently, we formulate the reconstruction of sharp features as a quadratic optimization problem. Our method's effectiveness has been validated through extensive testing using self-constructed models and 100 filleted models selected from the Fusion 360 Gallery dataset. The code for this paper is publicly available at https://github.com/xiaowuga/DeFillet.

Neural implicit representations are increasingly used to depict three-dimensional (3D) shapes owing to their inherent smoothness and compactness, contrasting with traditional discrete representations. Yet, the multilayer perceptron–based neural representation, because of its smooth nature, rounds sharp corners or edges, rendering it unsuitable for representing objects with sharp features like computer-aided design (CAD) models. Moreover, neural implicit representations need long training times to fit 3D shapes. While previous works address these issues separately, we present a unified neural implicit representation called Patch-Grid , which efficiently fits complex shapes, preserves sharp features delineating different patches, and can also represent surfaces with open boundaries and thin geometric features. Patch-Grid learns a signed distance field (SDF) to approximate an encompassing surface patch of the shape with a learnable patch feature volume. To form sharp edges and corners in a CAD model, Patch-Grid merges the learned SDFs via the constructive solid geometry (CSG) approach. Core to the merging process is a novel merge grid design that organizes different patch feature volumes in a common octree structure. This design choice ensures robust merging of multiple learned SDFs by confining the CSG operations to localized regions. Additionally, it drastically reduces the complexity of the CSG operations in each merging cell, allowing the proposed method to be trained in seconds to fit a complex shape at high fidelity. Experimental results demonstrate that the proposed Patch-Grid representation is capable of accurately reconstructing shapes with complex sharp features, open boundaries, and thin geometric elements, achieving state-of-the-art reconstruction quality with high computational efficiency within seconds.

Neural implicit shape representation has drawn significant attention in recent years due to its smoothness, differentiability, and topological flexibility. However, directly modeling the shape of a neural implicit surface, especially as the zero-level set of a neural signed distance function (SDF), with sparse geometric control is still a challenging task. Sparse input shape control typically includes 3D curve networks or, more generally, 3D curve sketches, which are unstructured and cannot be connected to form a curve network, and therefore more difficult to deal with. While 3D curve networks or curve sketches provide intuitive shape control, their sparsity and varied topology pose challenges in generating high-quality surfaces to meet such curve constraints. In this paper, we propose NeuVAS, a variational approach to shape modeling using neural implicit surfaces constrained under sparse input shape control, including unstructured 3D curve sketches as well as connected 3D curve networks. Specifically, we introduce a smoothness term based on a functional of surface curvatures to minimize shape variation of the zero-level set surface of a neural SDF. We also develop a new technique to faithfully model G 0 sharp feature curves as specified in the input curve sketches. Comprehensive comparisons with the state-of-the-art methods demonstrate the significant advantages of our method.

Man-made 3D shapes are naturally organized in parts and hierarchies; such structures provide important constraints for shape reconstruction and generation. Modeling shape structures is difficult, because there can be multiple hierarchies for a given shape, causing ambiguity, and across different categories, the shape structures are correlated with semantics, limiting generalization. We present StructRe , a structure rewriting system, as a novel approach to structured shape modeling. Given a 3D object represented by points and components, StructRe can rewrite it upward into more concise structures, or downward into more detailed structures; by iterating the rewriting process, hierarchies are obtained. Such a localized rewriting process enables probabilistic modeling of ambiguous structures and robust generalization across object categories. We train StructRe on PartNet data and show its generalization to cross-category and multiple object hierarchies, and test its extension to ShapeNet. We also demonstrate the benefits of probabilistic and generalizable structure modeling for shape reconstruction, generation and editing tasks.

Compressed representations of 3D shapes that are compact, accurate, and can be processed efficiently directly in compressed form, are extremely useful for digital media applications. Recent approaches in this space focus on learned implicit or parametric representations. While implicits are well suited for tasks such as in-out queries, they lack natural 2D parameterization, complicating tasks such as texture or normal mapping. Conversely, parametric representations support the latter tasks but are ill-suited for occupancy queries. We propose a novel learned alternative to these approaches, based on intersections of localized explicit , or height-field , surfaces. Since explicits can be trivially expressed both implicitly and parametrically, NESI directly supports a wider range of processing operations than implicit alternatives, including occupancy queries and parametric access. We represent input shapes using a collection of differently oriented height-field bounded half-spaces combined using volumetric Boolean intersections. We first tightly bound each input using a pair of oppositely oriented height-fields, forming a Double Height-Field (DHF) Hull . We refine this hull by intersecting it with additional localized height-fields (HFs) that capture surface regions in its interior. We minimize the number of HFs necessary to accurately capture each input and compactly encode both the DHF hull and the local HFs as neural functions defined over subdomains of \(\mathbb {R}^2\) . This reduced dimensionality encoding delivers high-quality compact approximations. Given similar parameter count, or storage capacity, NESI significantly reduces approximation error compared to the state-of-the-art, especially at lower parameter counts.

This paper introduces a method to synthesize a 3D tensor field within a constrained geometric domain represented as a tetrahedral mesh. Whereas previous techniques optimize for isotropic fields, we focus on anisotropic tensor fields that are smooth and aligned with the domain boundary or user guidance. The key ingredient of our method is a novel computational design framework, built on top of the symmetric orthogonally decomposable (odeco) tensor representation, to optimize the stretching ratios and orientations for each tensor in the domain. In contrast to past techniques designed only for isotropic tensors, we demonstrate the efficacy of our approach in generating smooth volumetric tensor fields with high anisotropy and shape conformity, especially for the domain with complex shapes. We apply these anisotropic tensor fields to various applications, such as anisotropic meshing, structural mechanics, and fabrication.

Color smudge operations from digital painting software enable users to create natural shading effects in high-fidelity paintings by interactively mixing colors. To precisely control results in traditional painting software, users tend to organize flat-filled color regions in multiple layers and smudge them to generate different color gradients. However, the requirement to carefully deal with regions makes the smudging process time-consuming and laborious, especially for non-professional users. This motivates us to investigate how to infer user-desired smudging effects when users smudge over regions in a single layer. To investigate improving color smudge performance, we first conduct a formative study. Following the findings of this study, we design SmartSmudge, a novel smudge tool that offers users dynamical smudge brushes and real-time region selection for easily generating natural and efficient shading effects. We demonstrate the efficiency and effectiveness of the proposed tool via a user study and quantitative analysis.

Despite recent advancements in the Large Reconstruction Model (LRM) demonstrating impressive results, when extending its input from single image to multiple images, it exhibits inefficiencies, subpar geometric and texture quality, as well as slower convergence speed than expected. It is attributed to that, LRM formulates 3D reconstruction as a naive images-to-3D translation problem, ignoring the strong 3D coherence among the input images. In this article, we propose a Multi-view Large Reconstruction Model (M-LRM) designed to reconstruct high-quality 3D shapes from multi-views in a 3D-aware manner. Specifically, we introduce a multi-view consistent cross-attention scheme to enable M-LRM to accurately query information from the input images. Moreover, we employ the 3D priors of the input multi-view images to initialize the triplane tokens. Compared to previous methods, the proposed M-LRM can generate 3D shapes of high fidelity. Experimental studies demonstrate that our model achieves a significant performance gain and faster training convergence.

Deducing the 3D face from a skull is a challenging task in forensic science and archaeology. This article proposes an end-to-end 3D face reconstruction pipeline and an exploration method that can conveniently create textured, realistic faces that match the given skull. To this end, we propose a tissue-guided face creation and adaptation scheme. With the help of the state-of-the-art text-to-image diffusion model and parametric face model, we first generate an initial reference 3D face, whose biological profile aligns with the given skull. Then, with the help of tissue thickness distribution, we modify these initial faces to match the skull through a latent optimization process. The joint distribution of tissue thickness is learned on a set of skull landmarks using a collection of scanned skull-face pairs. We also develop an efficient face adaptation tool to allow users to interactively adjust tissue thickness either globally or at local regions to explore different plausible faces. Experiments conducted on a real skull-face dataset demonstrated the effectiveness of our proposed pipeline in terms of reconstruction accuracy, diversity, and stability.

Neural implicit fields, such as the neural signed distance field (SDF) of a shape, have emerged as a powerful representation for many applications, e.g., encoding a 3D shape and performing collision detection. Typically, implicit fields are encoded by Multi-layer Perceptrons (MLP) with positional encoding (PE) to capture high-frequency geometric details. However, a notable side effect of such PE-equipped MLPs is the noisy artifacts present in the learned implicit fields. While increasing the sampling rate could in general mitigate these artifacts, in this paper we aim to explain this adverse phenomenon through the lens of Fourier analysis. We devise a tool to determine the appropriate sampling rate for learning an accurate neural implicit field without undesirable side effects. Specifically, we propose a simple yet effective method to estimate the intrinsic frequency of a given network with randomized weights based on the Fourier analysis of the network's responses. It is observed that a PE-equipped MLP has an intrinsic frequency much higher than the highest frequency component in the PE layer. Sampling against this intrinsic frequency following the Nyquist-Sannon sampling theorem allows us to determine an appropriate training sampling rate. We empirically show in the setting of SDF fitting that this recommended sampling rate is sufficient to secure accurate fitting results, while further increasing the sampling rate would not further noticeably reduce the fitting error. Training PE-equipped MLPs simply with our sampling strategy leads to performances superior to the existing methods.

Neural radiance fields (NeRF) have demonstrated impressive performance in novel view synthesis, but are still slow to render complex scenes at a high resolution. We introduce a novel method to boost the NeRF rendering speed by utilizing the temporal coherence between consecutive frames. Rather than computing features of each frame entirely from scratch, we reuse the coherent information (e.g., density and color) computed from the previous frames to help render the current frame, which significantly boosts rendering speed. To effectively manage the coherent information of previous frames, we introduce a history buffer with a multiple-plane structure, which is built online and updated from old frames to new frames. We name this buffer as multiple plane buffer (MPB). With this MPB, a new frame can be efficiently rendered using the warped features from previous frames. Extensive experiments on the NeRF-Synthetic, LLFF, and Mip-NeRF-360 datasets demonstrate that our method significantly boosts rendering efficiency and achieves 4× speedup on real-world scenes compared to the baseline methods while preserving competitive rendering quality.

Invisible orthodontics, commonly known as clear alignment treatment, offers a more comfortable and aesthetically pleasing alternative in orthodontic care, attracting considerable attention in the dental community in recent years. It replaces conventional metal braces with a series of removable, and transparent aligners. Each aligner is crafted to facilitate a gradual adjustment of the teeth, ensuring progressive stages of dental correction. This necessitates the design for teeth motion. Here we present an automatic method and a system for generating collision-free teeth motion planning while avoiding gaps between adjacent teeth, which is unacceptable in clinical practice. To tackle this task, we formulate it as a constrained optimization problem and utilize the interior point method for its solution. We also developed an interactive system that enables dentists to easily visualize and edit the paths. Our method significantly speeds up the clear aligner planning process, creating the desired motion paths for a full set of teeth in under five minutes-a task that typically requires several hours of manual work. Our experiments and user studies confirm the effectiveness of this method in planning teeth movement, showcasing its potential to streamline orthodontic procedures.

Structural representation is crucial for reconstructing and generating editable 3D shapes with part semantics. Recent 3D shape generation works employ complicated networks and structure definitions relying on hierarchical annotations and pay less attention to the details inside parts. In this paper, we propose the method that parameterizes the shared structure in the same category using a differentiable template and corresponding fixed-length parameters. Specific parameters are fed into the template to calculate cuboids that indicate a concrete shape. We utilize the boundaries of three-view renderings of each cuboid to further describe the inside details. Shapes are represented with the parameters and three-view details inside cuboids, from which the SDF can be calculated to recover the object. Benefiting from our fixed-length parameters and three-view details, our networks for reconstruction and generation are simple and effective to learn the latent space. Our method can reconstruct or generate diverse shapes with complicated details, and interpolate them smoothly. Extensive evaluations demonstrate the superiority of our method on reconstruction from point cloud, generation, and interpolation.

We propose a new method for computing smooth and integrable cross fields on 2D and 3D surfaces. our approach first computes smooth cross fields by minimizing the Dirichlet energy. Unlike existing optimization-based methods, our technique determines the singularity configuration-i.e., the number, locations, and indices of singularities-by iteratively adjusting them. Singularities can move, merge and split, akin to the behavior of like charges repelling and unlike charges attracting. Once all singularities stop moving, we obtain a cross field with (locally) the lowest Dirichlet energy. In simply connected domains, this cross field is guaranteed to be integrable. However, this property does not hold in multiply connected domains. To make a smooth cross field integrable, we construct a vector field $\bf c$c that characterizes the deviation of the cross field from a curl-free field. We then optimize the locations of singularities by moving them along the field lines of $\bf c$c. Our method is fundamentally different from existing integer programming-based approaches, as it avoids combinatorial optimization. It is fully automatic and includes a parameter to control the number of singularities. Our method is well suited for smooth models where exact boundary alignment and sparse hard directional constraints are desired, and can guide seamless conformal parameterization and T-junction-free quadrangulation.

While tooth alignment is crucial for digital dentistry, especially in orthodontic treatment, existing computer-aided methods mainly focus on the 3D dental crown but overlook the entire teeth, which is essential for applications in orthodontics. Besides, clinical orthodontic rules are not fully considered in these methods, i.e., there should be no collisions and gaps between teeth, the upper jaw and lower jaw should have correct occlusion relationships, the teeth should comply with a reasonable dental arch curve, etc. To generate optimal tooth alignment results, we propose a rule-based optimization method for solving the tooth alignment problem that takes into consideration the clinical rules functionally and aesthetically. We optimize rule-driven objective functions by adjusting the 6-DoF transformations of each tooth. Besides, our optimization formulation supports customization for different clinical scenarios by specifying the various energy terms. Extensive experiments, ablation studies, and user studies have been conducted to validate the effectiveness of our method. Quantitative and qualitative comparisons demonstrate that our method generates better tooth alignments than previous methods.

In this paper, we present WonderHuman to reconstruct dynamic human avatars from a monocular video for high-fidelity novel view synthesis. Previous dynamic human avatar reconstruction methods typically require the input video to have full coverage of the observed human body. However, in daily practice, one typically has access to limited viewpoints, such as monocular front-view videos, making it a cumbersome task for previous methods to reconstruct the unseen parts of the human avatar. To tackle the issue, we present WonderHuman, which leverages 2D generative diffusion model priors to achieve high-quality, photorealistic reconstructions of dynamic human avatars from monocular videos, including accurate rendering of unseen body parts. Our approach introduces a Dual-Space Optimization technique, applying Score Distillation Sampling (SDS) in both canonical and observation spaces to ensure visual consistency and enhance realism in dynamic human reconstruction. Additionally, we present a View Selection strategy and Pose Feature Injection to enforce the consistency between SDS predictions and observed data, ensuring pose-dependent effects and higher fidelity in the reconstructed avatar. In the experiments, our method achieves SOTA performance in producing photorealistic renderings from the given monocular video, particularly for those challenging unseen parts.

Extraction of a high-fidelity 3D medial axis is a crucial operation in CAD. When dealing with a polygonal model as input, ensuring accuracy and tidiness becomes challenging due to discretization errors inherent in the mesh surface. Commonly, existing approaches yield medial-axis surfaces with various artifacts, including zigzag boundaries, bumpy surfaces, unwanted spikes, and non-smooth stitching curves. Considering that the surface of a CAD model can be easily decomposed into a collection of surface patches, its 3D medial axis can be extracted by computing the Voronoi diagram of these surface patches, where each surface patch serves as a generator. However, no solver currently exists for accurately computing such an extended Voronoi diagram. Under the assumption that each generator defines a linear distance field over a sufficiently small range, our approach operates by tetrahedralizing the region of interest and computing the medial axis within each tetrahedral element. Just as SurfaceVoronoi computes surface-based Voronoi diagrams by cutting a 3D prism with 3D planes (each plane encodes a linear field in a triangle), the key operation in this paper is to conduct the hyperplane cutting process in 4D, where each hyperplane encodes a linear field in a tetrahedron. In comparison with the state-of-the-art, our algorithm produces better outcomes. Furthermore, it can also be used to compute the offset surface.

Nowadays, orthodontics has become an important part of modern personal life to assist one in improving mastication and raising self-esteem. However, the quality of orthodontic treatment still heavily relies on the empirical evaluation of experienced doctors, which lacks quantitative assessment and requires patients to visit clinics frequently for in-person examination. To resolve the aforementioned problem, we propose a novel and practical mobile device-based framework for precisely measuring tooth movement in treatment, so as to simplify and strengthen the traditional tooth monitoring process. To this end, we formulate the tooth movement monitoring task as a multi-view multi-object pose estimation problem via different views that capture multiple texture-less and severely occluded objects (i.e. teeth). Specifically, we exploit a pre-scanned 3D tooth model and a sparse set of multi-view tooth images as inputs for our proposed tooth monitoring framework. After extracting tooth contours and localizing the initial camera pose of each view from the initial configuration, we propose a joint pose estimation scheme to precisely estimate the 3D pose of each individual tooth, so as to infer their relative offsets during treatment. Furthermore, we introduce the metric of Relative Pose Bias to evaluate the individual tooth pose accuracy in a small scale. We demonstrate that our approach is capable of reaching high accuracy and efficiency as practical orthodontic treatment monitoring requires.

The reconstruction of indoor scenes from multi-view RGB images is challenging due to the coexistence of flat and texture-less regions alongside delicate and fine-grained regions. Recent methods leverage neural radiance fields aided by predicted surface normal priors to recover the scene geometry. These methods excel in producing complete and smooth results for floor and wall areas. However, they struggle to capture complex surfaces with high-frequency structures due to the inadequate neural representation and the inaccurately predicted normal priors. This work aims to reconstruct high-fidelity surfaces with fine-grained details by addressing the above limitations. To improve the capacity of the implicit representation, we propose a hybrid architecture to represent low-frequency and high-frequency regions separately. To enhance the normal priors, we introduce a simple yet effective image sharpening and denoising technique, coupled with a network that estimates the pixel-wise uncertainty of the predicted surface normal vectors. Identifying such uncertainty can prevent our model from being misled by unreliable surface normal supervisions that hinder the accurate reconstruction of intricate geometries. Experiments on the benchmark datasets show that our method outperforms existing methods in terms of reconstruction quality. Furthermore, the proposed method also generalizes well to real-world indoor scenarios captured by our hand-held mobile phones.

Tooth motion generation is an essential task in digital orthodontic treatment for precise and quick dental healthcare, which aims to generate the whole intermediate tooth motion process given the initial pathological and target ideal tooth alignments. Most prior works for multi-agent motion planning problems usually result in complex solutions. Moreover, the occlusal relationship between upper and lower teeth is often overlooked. In this paper, we propose a collaborative tooth motion diffusion model. The critical insight is to remodel the problem as a diffusion process. In this sense, we model the whole tooth motion distribution with a diffusion model and transform the planning problem into a sampling process from this distribution. We design a tooth latent representation to provide accurate conditional guides consisting of two key components: the tooth frame represents the position and posture, and the tooth latent shape code represents the geometric morphology. Subsequently, we present a collaborative diffusion model to learn the multi-tooth motion distribution based on inter-tooth and occlusal constraints, which are implemented by graph structure and new loss functions, respectively. Extensive qualitative and quantitative experiments demonstrate the superiority of our framework in the application of orthodontics compared with state-of-the-art methods.

Adversarial attacks on 3D point clouds often exhibit unsatisfactory imperceptibility, which primarily stems from the disregard for manifold-aware distortion, i.e., distortion of the underlying 2-manifold surfaces. In this paper, we develop novel manifold constraints to reduce such distortion, aiming to enhance the imperceptibility of adversarial attacks on 3D point clouds. Specifically, we construct a bijective manifold mapping between point clouds and a simple parameter shape using an invertible auto-encoder. Consequently, manifold-aware distortion during attacks can be captured within the parameter space. By enforcing manifold constraints that preserve local properties of the parameter shape, manifold-aware distortion is effectively mitigated, ultimately leading to enhanced imperceptibility. Extensive experiments demonstrate that integrating manifold constraints into conventional adversarial attack solutions yields superior imperceptibility, outperforming the state-of-the-art methods.

The advent of neural 3D Gaussians [21] has recently brought about a revolution in the field of neural rendering, facilitating the generation of high-quality renderings at real-time speeds. However, the explicit and discrete repre-sentation encounters challenges when applied to scenes fea-turing reflective surfaces. In this paper, we present Gaus-sian Shader, a novel method that applies a simplified shading function on 3D Gaussians to enhance the neural ren-dering in scenes with reflective surfaces while preserving the training and rendering efficiency. The main challenge in applying the shading function lies in the accurate nor-mal estimation on discrete 3D Gaussians. Specifically, we proposed a novel normal estimation framework based on the shortest axis directions of 3D Gaussians with a deli-cately designed loss to make the consistency between the normals and the geometries of Gaussian spheres. Experiments show that GaussianShader strikes a commendable balance between efficiency and visual quality. Our method surpasses Gaussian Splatting [21] in PSNR on specular object datasets, exhibiting an improvement of 1.57dB. When compared to prior works handling reflective surfaces, such as Ref-NeRF [45], our optimization time is significantly accelerated (23h vs. 0.58h). Please click on our project web-site to see more results

In this work, we introduce Wonder3D, a novel method for efficiently generating high-fidelity textured meshes from single-view images. Recent methods based on Score Distillation Sampling (SDS) have shown the potential to recover 3D geometry from 2D diffusion priors, but they typically suffer from time-consuming per-shape optimization and inconsistent geometry. In contrast, certain works di-rectly produce 3D information via fast network inferences, but their results are often of low quality and lack geometric details. To holistically improve the quality, consistency, and efficiency of single-view reconstruction tasks, we pro-pose a cross-domain diffusion model that generates multi-view normal maps and the corresponding color images. To ensure the consistency of generation, we employ a multi-view cross-domain attention mechanism that facilitates information exchange across views and modalities. Lastly, we introduce a geometry-aware normal fusion algorithm that extracts high-quality surfaces from the multi-view 2D representations in only 2 r-;» 3 minutes. Our extensive evaluations demonstrate that our method achieves high-quality reconstruction results, robust generalization, and good efficiency compared to prior works.

Deep neural networks (DNNs) often display overconfidence when encountering out-of-distribution (OOD) samples, posing significant challenges in real-world applications. Capitalizing on the observation that responses on convolutional kernels are generally more pronounced for in-distribution (ID) samples than for OOD ones, this paper proposes the COnvolutional REsponse-based Score (CORES) to exploit these discrepancies for OOD detection. Initially, CORES delves into the extremities of convolutional responses by considering both their magnitude and the frequency of significant values. Moreover, through backtracking from the most prominent predictions, CORES effectively pinpoints sample-relevant kernels across different layers. These kernels, which exhibit a strong correlation to input samples, are integral to CORES's OOD detection capability. Comprehensive experiments across various ID and OOD settings demonstrate CORES's effectiveness in OOD detection and its superiority to the state-of-the-art methods.

A diffusion model, which is formulated to produce an image using thousands of denoising steps, usually suffers from a slow inference speed. Existing acceleration algorithms simplify the sampling by skipping most steps yet exhibit considerable performance degradation. By viewing the generation of diffusion models as a discretized integral process, we argue that the quality drop is partly caused by applying an inaccurate integral direction to a timestep interval. To rectify this issue, we propose a timestep tuner that helps find a more accurate integral direction for a particular interval at the minimum cost. Specifically, at each denoising step, we replace the original parameterization by conditioning the network on a new timestep, enforcing the sampling distribution towards the real one. Extensive experiments show that our plug-in design can be trained efficiently and boost the inference performance of various state-of-the-art acceleration methods, especially when there are few denoising steps. For example, when using 10 denoising steps on LSUN Bedroom dataset, we improve the FID of DDIM from 9.65 to 6.07, simply by adopting our method for a more appropriate set of timesteps. Code is available at https://github.com/THU-LYJ-Lab/time-tuner.

The field of 3D detailed human mesh reconstruction has made significant progress in recent years. However, current methods still face challenges when used in industrial applications due to unstable results, low-quality meshes, and a lack of UV unwrapping and skinning weights. In this paper, we present SHERT, a novel pipeline that can reconstruct semantic human meshes with textures and high-precision details. SHERT applies semantic- and normal-based sampling between the detailed surface (e.g. mesh and SDF) and the corresponding SMPL-X model to obtain a partially sampled semantic mesh and then generates the complete semantic mesh by our specifically designed self-supervised completion and refinement networks. Using the complete semantic mesh as a basis, we employ a texture diffusion model to create human textures that are driven by both images and texts. Our reconstructed meshes have stable UV unwrapping, high-quality triangle meshes, and consistent semantic information. The given SMPL-X model provides semantic information and shape priors, allowing SHERT to perform well even with incorrect and incomplete inputs. The semantic information also makes it easy to substitute and animate different body parts such as the face, body, and hands. Quantitative and qualitative experiments demonstrate that SHERT is capable of producing high-fidelity and robust semantic meshes that outperform state-of-the-art methods.

Matching cross-modality features between images and point clouds is a fundamental problem for image-to-point cloud registration. However, due to the modality difference between images and points, it is difficult to learn robust and discriminative cross-modality features by existing metric learning methods for feature matching. Instead of applying metric learning on cross-modality data, we propose to unify the modality between images and point clouds by pretrained large-scale models first, and then establish robust correspondence within the same modality. We show that the intermediate features, called diffusion features, extracted by depth-to-image diffusion models are semantically consistent between images and point clouds, which enables the building of coarse but robust cross-modality correspondences. We further extract geometric features on depth maps produced by the monocular depth estimator. By matching such geometric features, we significantly improve the accuracy of the coarse correspondences produced by diffusion features. Extensive experiments demonstrate that without any task-specific training, direct utilization of both features produces accurate image-to-point cloud registration. On three public indoor and outdoor benchmarks, the proposed method averagely achieves a 20.6 percent improvement in Inlier Ratio, a $3.0\times$ higher Inlier Number, and a 48.6 percent improvement in Registration Recall than existing state-of-the-arts. The code and additional results are available at \url{https://whu-usi3dv.github.io/FreeReg/}.

We propose a Pose-Free Large Reconstruction Model (PF-LRM) for reconstructing a 3D object from a few unposed images even with little visual overlap, while simultaneously estimating the relative camera poses in ~1.3 seconds on a single A100 GPU. PF-LRM is a highly scalable method utilizing self-attention blocks to exchange information between 3D object tokens and 2D image tokens; we predict a coarse point cloud for each view, and then use a differentiable Perspective-n-Point (PnP) solver to obtain camera poses. When trained on a huge amount of multi-view posed data of ~1M objects, PF-LRM shows strong cross-dataset generalization ability, and outperforms baseline methods by a large margin in terms of pose prediction accuracy and 3D reconstruction quality on various unseen evaluation datasets. We also demonstrate our model's applicability in downstream text/image-to-3D task with fast feed-forward inference. Our project website is at: https://totoro97.github.io/pf-lrm.

In this paper, we present a novel diffusion model called SyncDreamer that generates multiview-consistent images from a single-view image. Using pretrained large-scale 2D diffusion models, recent work Zero123 demonstrates the ability to generate plausible novel views from a single-view image of an object. However, maintaining consistency in geometry and colors for the generated images remains a challenge. To address this issue, we propose a synchronized multiview diffusion model that models the joint probability distribution of multiview images, enabling the generation of multiview-consistent images in a single reverse process. SyncDreamer synchronizes the intermediate states of all the generated images at every step of the reverse process through a 3D-aware feature attention mechanism that correlates the corresponding features across different views. Experiments show that SyncDreamer generates images with high consistency across different views, thus making it well-suited for various 3D generation tasks such as novel-view-synthesis, text-to-3D, and image-to-3D. Project page: https://liuyuan-pal.github.io/SyncDreamer/.

3D Gaussian Splatting (3DGS) has recently revolutionized the field of neural rendering with its high fidelity and efficiency. However, 3DGS heavily depends on the initialized point cloud produced by Structure-from-Motion (SfM) techniques. When tackling large-scale scenes that unavoidably contain texture-less surfaces, SfM techniques fail to produce enough points in these surfaces and cannot provide good initialization for 3DGS. As a result, 3DGS suffers from difficult optimization and low-quality renderings. In this paper, inspired by classic multi-view stereo (MVS) techniques, we propose GaussianPro, a novel method that applies a progressive propagation strategy to guide the densification of the 3D Gaussians. Compared to the simple split and clone strategies used in 3DGS, our method leverages the priors of the existing reconstructed geometries of the scene and utilizes patch matching to produce new Gaussians with accurate positions and orientations. Experiments on both large-scale and small-scale scenes validate the effectiveness of our method. Our method significantly surpasses 3DGS on the Waymo dataset, exhibiting an improvement of 1.15dB in terms of PSNR. Codes and data are available at https://github.com/kcheng1021/GaussianPro.

Generative adversarial networks (GANs) usually struggle in learning from highly diverse data, whose underlying manifold is complex. In this work, we revisit the mathematical foundations of GANs, and theoretically reveal that the native adversarial loss for GAN training is insufficient to fix the problem of $\textit{subsets with positive Lebesgue measure of the generated data manifold lying out of the real data manifold}$. Instead, we find that score matching serves as a promising solution to this issue thanks to its capability of persistently pushing the generated data points towards the real data manifold. We thereby propose to improve the optimization of GANs with score matching regularity (SMaRt). Regarding the empirical evidences, we first design a toy example to show that training GANs by the aid of a ground-truth score function can help reproduce the real data distribution more accurately, and then confirm that our approach can consistently boost the synthesis performance of various state-of-the-art GANs on real-world datasets with pre-trained diffusion models acting as the approximate score function. For instance, when training Aurora on the ImageNet $64\times64$ dataset, we manage to improve FID from 8.87 to 7.11, on par with the performance of one-step consistency model. Code is available at https://github.com/thuxmf/SMaRt.

Depth estimation is a crucial task in computer vision, applicable to various domains such as 3D reconstruction, robotics, and autonomous driving. In particular, thermal-based depth estimation has unique advantages, including night-time vision. However, the existing depth estimation method remains challenging in robust generalization due to limited data resources and spectral differences between thermal and RGB images. In this paper, we present a self-supervised approach to enhance thermal-based depth estimation by leveraging pre-trained visual models initially designed for RGB data. In detail, we design a novel two-stage training strategy, incorporating Low-rank Adapters and Convolutional Adapters, which not only significantly improves accuracy and robustness but also enables impressive zero-shot generalization capabilities. Our method outperforms existing thermal-based depth estimation models, opening new possibilities for cross-modal applications in computer vision and robotics research.

This paper presents a flexible representation of neural radiance fields based on multi-plane images (MPI), for high-quality view synthesis of complex scenes. MPI with Normalized Device Coordinate (NDC) parameterization is widely used in NeRF learning for its simple definition, easy calculation, and powerful ability to represent unbounded scenes. However, existing NeRF works that adopt MPI representation for novel view synthesis can only handle simple forward-facing unbounded scenes (e.g., the scenes in the LLFF dataset), where the input cameras are all observing in similar directions with small relative translations. Hence, extending these MPIbased methods to more complex scenes like large-range or even 360-degree scenes is very challenging. In this paper, we explore the potential of MPI and show that MPI can synthesize high-quality novel views of complex scenes with diverse camera distributions and view directions, which are not only limited to simple forward-facing scenes. Our key idea is to encode the neural radiance field with multiple MPIs facing different directions and blend them with an adaptive blending operation. For each region of the scene, the blending operation gives larger blending weights to those advantaged MPIs with stronger local representation abilities while giving lower weights to those with weaker representation abilities. Such blending operation automatically modulates the multiple MPIs to appropriately represent the diverse local density and color information. Experiments on the KITTI dataset and ScanNet dataset demonstrate that our proposed MMPI synthesizes high-quality images from diverse camera pose distributions and is fast to train, outperforming the previous fast-training NeRF methods for novel view synthesis. Moreover, we show that MMPI can encode extremely long trajectories and produce novel view renderings, demonstrating its potential in applications like autonomous driving. Our demo video is available at https://youtube.com/watch?v=mbNKwN5urC8.

Adversarial attacks on point clouds are crucial for assessing and improving the adversarial robustness of 3D deep learning models. Despite leveraging various geometric constraints, current adversarial attack strategies often suffer from inadequate imperceptibility. Given that adversarial perturbations tend to disrupt the inherent symmetry in objects, we recognize this disruption as the primary cause of the lack of imperceptibility in these attacks. In this paper, we introduce a novel framework, symmetry-aware imperceptible adversarial attacks on 3D point clouds (SymAttack), to address this issue. Our approach starts by identifying part- and patch-level symmetry elements, and grouping points based on semantic and Euclidean distances, respectively. During the adversarial attack iterations, we intentionally adjust the perturbation vectors on symmetric points relative to their symmetry plane. By preserving symmetry within the attack process, SymAttack significantly enhances imperceptibility. Extensive experiments validate the effectiveness of SymAttack in generating imperceptible adversarial point clouds, demonstrating its superiority over the state-of-the-art methods.

In this paper, we introduce **Era3D**, a novel multiview diffusion method that generates high-resolution multiview images from a single-view image. Despite significant advancements in multiview generation, existing methods still suffer from camera prior mismatch, inefficacy, and low resolution, resulting in poor-quality multiview images. Specifically, these methods assume that the input images should comply with a predefined camera type, e.g. a perspective camera with a fixed focal length, leading to distorted shapes when the assumption fails. Moreover, the full-image or dense multiview attention they employ leads to a dramatic explosion of computational complexity as image resolution increases, resulting in prohibitively expensive training costs. To bridge the gap between assumption and reality, Era3D first proposes a diffusion-based camera prediction module to estimate the focal length and elevation of the input image, which allows our method to generate images without shape distortions. Furthermore, a simple but efficient attention layer, named row-wise attention, is used to enforce epipolar priors in the multiview diffusion, facilitating efficient cross-view information fusion. Consequently, compared with state-of-the-art methods, Era3D generates high-quality multiview images with up to a 512×512 resolution while reducing computation complexity of multiview attention by 12x times. Comprehensive experiments demonstrate the superior generation power of Era3D- it can reconstruct high-quality and detailed 3D meshes from diverse single-view input images, significantly outperforming baseline multiview diffusion methods.

Surface parameterization plays an essential role in numerous computer graphics and geometry processing applications. Traditional parameterization approaches are designed for high-quality meshes laboriously created by specialized 3D modelers, thus unable to meet the processing demand for the current explosion of ordinary 3D data. Moreover, their working mechanisms are typically restricted to certain simple topologies, thus relying on cumbersome manual efforts (e.g., surface cutting, part segmentation) for pre-processing. In this paper, we introduce the Flatten Anything Model (FAM), an unsupervised neural architecture to achieve global free-boundary surface parameterization via learning point-wise mappings between 3D points on the target geometric surface and adaptively-deformed UV coordinates within the 2D parameter domain. To mimic the actual physical procedures, we ingeniously construct geometrically-interpretable sub-networks with specific functionalities of surface cutting, UV deforming, unwrapping, and wrapping, which are assembled into a bi-directional cycle mapping framework. Compared with previous methods, our FAM directly operates on discrete surface points without utilizing connectivity information, thus significantly reducing the strict requirements for mesh quality and even applicable to unstructured point cloud data. More importantly, our FAM is fully-automated without the need for pre-cutting and can deal with highly-complex topologies, since its learning process adaptively finds reasonable cutting seams and UV boundaries. Extensive experiments demonstrate the universality, superiority, and inspiring potential of our proposed neural surface parameterization paradigm. Our code is available at https://github.com/keeganhk/FlattenAnything.

Given the rapid advancements in geometric deep-learning techniques, there has been a dedicated effort to create mesh-based convolutional operators that act as a link between irregular mesh structures and widely adopted backbone networks . Despite the numerous advantages of Convolutional Neural Networks (CNNs) over Multi-Layer Perceptrons (MLPs), mesh-oriented CNNs often require intricate network architectures to tackle irregularities of a triangular mesh. These architectures not only demand that the mesh be manifold and watertight but also impose constraints on the abundance of training samples . In this paper, we note that for specific tasks such as mesh classification and semantic part segmentation, large-scale shape features play a pivotal role . This is in contrast to the realm of shape correspondence, where a comprehensive understanding of 3D shapes necessitates considering both local and global characteristics. Inspired by this key observation, we introduce a task-driven neural network architecture that seamlessly operates in an end-to-end fashion. Our method takes as input mesh vertices equipped with the heat kernel signature (HKS) and dihedral angles between adjacent faces . Notably, we replace the conventional convolutional module, commonly found in ResNet architectures, with MLPs and incorporate Layer Normalization (LN) to facilitate layer-wise normalization. Our approach, with a seemingly straightforward network architecture, demonstrates an accuracy advantage. It exhibits a marginal 0.1% improvement in the mesh classification task and a substantial 1.8% enhancement in the mesh part segmentation task compared to state-of-the-art methodologies. Moreover, as the number of training samples decreases to 1/50 or even 1/100, the accuracy advantage of our approach becomes more pronounced. In summary, our convolution-free network is tailored for specific tasks relying on large-scale shape features and excels in the situation with a limited number of training samples, setting itself apart from state-of-the-art methodologies.

Abstract We introduce Coverage Axis++, a novel and efficient approach to 3D shape skeletonization. The current state‐of‐the‐art approaches for this task often rely on the watertightness of the input [LWS*15; PWG*19; PWG*19] or suffer from substantial computational costs [DLX*22; CD23], thereby limiting their practicality. To address this challenge, Coverage Axis++ proposes a heuristic algorithm to select skeletal points, offering a high‐accuracy approximation of the Medial Axis Transform (MAT) while significantly mitigating computational intensity for various shape representations. We introduce a simple yet effective strategy that considers shape coverage, uniformity, and centrality to derive skeletal points. The selection procedure enforces consistency with the shape structure while favoring the dominant medial balls, which thus introduces a compact underlying shape representation in terms of MAT. As a result, Coverage Axis++ allows for skeletonization for various shape representations (e.g., water‐tight meshes, triangle soups, point clouds), specification of the number of skeletal points, few hyperparameters, and highly efficient computation with improved reconstruction accuracy. Extensive experiments across a wide range of 3D shapes validate the efficiency and effectiveness of Coverage Axis++. Our codes are available at https://github.com/Frank-ZY-Dou/Coverage_Axis .

the Best Paper Award Committee to select the Best Paper.After careful deliberation, the following paper was chosen with the unanimous consensus as the winner, on the basis of its intellectual merit and potential impact:Visual attention network [1] Two other papers were awarded an

We introduce a novel approach to learn geometries such as depth and surface normal from images while incorporating geometric context. The difficulty of reliably capturing geometric context in existing methods impedes their ability to accurately enforce the consistency between the different geometric properties, thereby leading to a bottleneck of geometric estimation quality. We therefore propose the Adaptive Surface Normal (ASN) constraint, a simple yet efficient method. Our approach extracts geometric context that encodes the geometric variations present in the input image and correlates depth estimation with geometric constraints. By dynamically determining reliable local geometry from randomly sampled candidates, we establish a surface normal constraint, where the validity of these candidates is evaluated using the geometric context. Furthermore, our normal estimation leverages the geometric context to prioritize regions that exhibit significant geometric variations, which makes the predicted normals accurately capture intricate and detailed geometric information. Through the integration of geometric context, our method unifies depth and surface normal estimations within a cohesive framework, which enables the generation of high-quality 3D geometry from images. We validate the superiority of our approach over state-of-the-art methods through extensive evaluations and comparisons on diverse indoor and outdoor datasets, showcasing its efficiency and robustness.

Neural Radiance Field (NeRF) has achieved substantial progress in novel view synthesis given multi-view images. Recently, some works have attempted to train a NeRF from a single image with 3D priors. They mainly focus on a limited field of view with a few occlusions, which greatly limits their scalability to real-world 360-degree panoramic scenarios with large-size occlusions. In this paper, we present PERF, a 360-degree novel view synthesis framework that trains a panoramic neural radiance field from a single panorama. Notably, PERF allows 3D roaming in a complex scene without expensive and tedious image collection. To achieve this goal, we propose a novel collaborative RGBD inpainting method and a progressive inpainting-and-erasing method to lift up a 360-degree 2D scene to a 3D scene. Specifically, we first predict a panoramic depth map as initialization given a single panorama and reconstruct visible 3D regions with volume rendering. Then we introduce a collaborative RGBD inpainting approach into a NeRF for completing RGB images and depth maps from random views, which is derived from an RGB Stable Diffusion model and a monocular depth estimator. Finally, we introduce an inpainting-and-erasing strategy to avoid inconsistent geometry between a newly-sampled view and reference views. The two components are integrated into the learning of NeRFs in a unified optimization framework and achieve promising results. Extensive experiments on Replica and a new dataset PERF-in-the-wild demonstrate the superiority of our PERF over state-of-the-art methods. Our PERF can be widely used for real-world applications, such as panorama-to-3D, text-to-3D, and 3D scene stylization applications.

Applying diffusion models to image-to-image translation (I2I) has recently received increasing attention due to its practical applications. Previous attempts inject information from the source image into each denoising step for an iterative refinement, thus resulting in a time-consuming implementation. We propose an efficient method that equips a diffusion model with a lightweight translator, dubbed a Diffusion Model Translator (DMT), to accomplish I2I. Specifically, we first offer theoretical justification that in employing the pioneering DDPM work for the I2I task, it is both feasible and sufficient to transfer the distribution from one domain to another only at some intermediate step. We further observe that the translation performance highly depends on the chosen timestep for domain transfer, and therefore propose a practical strategy to automatically select an appropriate timestep for a given task. We evaluate our approach on a range of I2I applications, including image stylization, image colorization, segmentation to image, and sketch to image, to validate its efficacy and general utility. The comparisons show that our DMT surpasses existing methods in both quality and efficiency. Code is available at https://github.com/THU-LYJ-Lab/dmt.

Dynamic 3D point cloud sequences serve as one of the most common and practical representation modalities of dynamic real-world environments. However, their unstructured nature in both spatial and temporal domains poses significant challenges to effective and efficient processing. Existing deep point cloud sequence modeling approaches imitate the mature 2D video learning mechanisms by developing complex spatio-temporal point neighbor grouping and feature aggregation schemes, often resulting in methods lacking effectiveness, efficiency, and expressive power. In this paper, we propose a novel generic representation called Structured Point Cloud Videos (SPCVs). Intuitively, by leveraging the fact that 3D geometric shapes are essentially 2D manifolds, SPCV re-organizes a point cloud sequence as a 2D video with spatial smoothness and temporal consistency, where the pixel values correspond to the 3D coordinates of points. The structured nature of our SPCV representation allows for the seamless adaptation of well-established 2D image/video techniques, enabling efficient and effective processing and analysis of 3D point cloud sequences. To achieve such re-organization, we design a self-supervised learning pipeline that is geometrically regularized and driven by self-reconstructive and deformation field learning objectives. Additionally, we construct SPCV-based frameworks for both low-level and high-level 3D point cloud sequence processing and analysis tasks, including action recognition, temporal interpolation, and compression. Extensive experiments demonstrate the versatility and superiority of the proposed SPCV, which has the potential to offer new possibilities for deep learning on unstructured 3D point cloud sequences.

Teleoperation systems find many applications from earlier search-and-rescue to more recent daily tasks. It is widely acknowledged that using external sensors can decouple the view of the remote scene from the motion of the robot arm during manipulation, facilitating the control task. However, this design requires the coordination of multiple operators or may exhaust a single operator as s/he needs to control both the manipulator arm and the external sensors. To address this challenge, our work introduces a viewpoint prediction model, the first data-driven approach that autonomously adjusts the viewpoint of a dynamic camera to assist in telemanipulation tasks. This model is parameterized by a deep neural network and trained on a set of human demonstrations. We propose a contrastive learning scheme that leverages viewpoints in a camera trajectory as contrastive data for network training. We demonstrated the effectiveness of the proposed viewpoint prediction model by integrating it into a real-world robotic system for telemanipulation. User studies reveal that our model outperforms several camera control methods in terms of control experience and reduces the perceived task load compared to manual camera control. As an assistive module of a telemanipulation system, our method significantly reduces task completion time for users who choose to adopt its recommendation.

Bimodal objects, such as the checkerboard pattern used in camera calibration, markers for object tracking, and text on road signs, to name a few, are prevalent in our daily lives and serve as a visual form to embed information that can be easily recognized by vision systems. While binarization from intensity images is crucial for extracting the embedded information in the bimodal objects, few previous works consider the task of binarization of blurry images due to the relative motion between the vision sensor and the environment. The blurry images can result in a loss in the binarization quality and thus degrade the downstream applications where the vision system is in motion. Recently, neuromorphic cameras offer new capabilities for alleviating motion blur, but it is non-trivial to first deblur and then binarize the images in a real-time manner. In this work, we propose an event-based binary reconstruction method that leverages the prior knowledge of the bimodal target's properties to perform inference independently in both event space and image space and merge the results from both domains to generate a sharp binary image. We also develop an efficient integration method to propagate this binary image to high frame rate binary video. Finally, we develop a novel method to naturally fuse events and images for unsupervised threshold identification. The proposed method is evaluated in publicly available and our collected data sequence, and shows the proposed method can outperform the SOTA methods to generate high frame rate binary video in real-time on CPU-only devices.

Recent progress in blind face restoration has resulted in producing high-quality restored results for static images. However, efforts to extend these advancements to video scenarios have been minimal, partly because of the absence of benchmarks that allow for a comprehensive and fair comparison. In this work, we first present a fair evaluation benchmark, in which we first introduce a Real-world Low-Quality Face Video benchmark (RFV-LQ), evaluate several leading image-based face restoration algorithms, and conduct a thorough systematical analysis of the benefits and challenges associated with extending blind face image restoration algorithms to degraded face videos. Our analysis identifies several key issues, primarily categorized into two aspects: significant jitters in facial components and noise-shape flickering between frames. To address these issues, we propose a Temporal Consistency Network (TCN) cooperated with alignment smoothing to reduce jitters and flickers in restored videos. TCN is a flexible component that can be seamlessly plugged into the most advanced face image restoration algorithms, ensuring the quality of image-based restoration is maintained as closely as possible. Extensive experiments have been conducted to evaluate the effectiveness and efficiency of our proposed TCN and alignment smoothing operation.

With the fast development of the 5G wireless communications, the Internet of Things (IoT) becomes a hot research topic. Unmanned aerial vehicles (UAVs), due to the high mobility and low labor cost, have a big potential to be applied in the future IoT communication networks, e.g., data collection in remote areas. In this paper, we take a UAV as a monitor and an IoT device as an agent, and study how to utilize UAVs to establish network coverage and enhance the overall performance. Given ground agents and aerial monitors, each agent needs to be supervised by one monitor that can at most take charge of certain amount, while such assignment should guarantee the required service quality. This is much different from the conventional assumption that each monitor owns exactly a fixed number of agents without considering sensing quality under limited transmit power. Suppose that a monitor supervises an agent with a cost as the negative value of transmission rate in Rician fading, we then maximize the sum of transmission induced by every monitor-agent connection constrained with workload capacity for each monitor. To achieve the above goals, we first present a fast algorithm to report the least-cost assignment plan. Then, we seek for the minimum number of monitors to maintain the required service quality. Last, we discuss the assignment problem in the scenario of dynamic agents and dynamic monitors. We also give a set of strategies on how to initialize assignment, optimize monitor locations and manage power consumption. Extensive experimental results on both simulated datasets and real-life traffic data demonstrate our effectiveness and high performance.

Despite recent advances in reconstructing an organic model with the neural signed distance function (SDF), the high-fidelity reconstruction of a CAD model directly from low-quality unoriented point clouds remains a significant challenge. In this paper, we address this challenge based on the prior observation that the surface of a CAD model is generally composed of piecewise surface patches, each approximately developable even around the feature line. Our approach, named NeurCADRecon , is self-supervised, and its loss includes a developability term to encourage the Gaussian curvature toward 0 while ensuring fidelity to the input points (see the teaser figure). Noticing that the Gaussian curvature is non-zero at tip points, we introduce a double-trough curve to tolerate the existence of these tip points. Furthermore, we develop a dynamic sampling strategy to deal with situations where the given points are incomplete or too sparse. Since our resulting neural SDFs can clearly manifest sharp feature points/lines, one can easily extract the feature-aligned triangle mesh from the SDF and then decompose it into smooth surface patches, greatly reducing the difficulty of recovering the parametric CAD design. A comprehensive comparison with existing state-of-the-art methods shows the significant advantage of our approach in reconstructing faithful CAD shapes.

We present a novel topology-preserving 3D medial axis computation framework based on volumetric restricted power diagram (RPD), while preserving the medial features and geometric convergence simultaneously, for both 3D CAD and organic shapes. The volumetric RPD discretizes the input 3D volume into sub-regions given a set of medial spheres. With this intermediate structure, we convert the homotopy equivalency between the generated medial mesh and the input 3D shape into a localized contractibility checking for each restricted element (power cell, power face, power edge), by checking their connected components and Euler characteristics. We further propose a fractional Euler characteristic algorithm for efficient GPU-based computation of Euler characteristic for each restricted element on the fly while computing the volumetric RPD. Compared with existing voxel-based or point-cloud-based methods, our approach is the first to adaptively and directly revise the medial mesh without globally modifying the dependent structure, such as voxel size or sampling density, while preserving its topology and medial features. In comparison with the feature preservation method MATFP [Wang et al. 2022], our method provides geometrically comparable results with fewer spheres and more robustly captures the topology of the input 3D shape.

Surface offsetting is a crucial operation in digital geometry processing and computer-aided design, where an offset is defined as an iso-value surface of the distance field. A challenge emerges as even smooth surfaces can exhibit sharp features in their offsets due to the non-differentiable characteristics of the underlying distance field. Prevailing approaches to the offsetting problem involve approximating the distance field and then extracting the iso-surface. However, even with dual contouring (DC), there is a risk of degrading sharp feature points/lines due to the inaccurate discretization of the distance field. This issue is exacerbated when the input is a piecewise-linear triangle mesh. This study is inspired by the observation that a triangle-based distance field, unlike the complex distance field rooted at the entire surface, remains smooth across the entire 3D space except at the triangle itself. With a polygonal surface comprising n triangles, the final distance field for accommodating the offset surface is determined by minimizing these n triangle-based distance fields. In implementation, our approach starts by tetrahedralizing the space around the offset surface, enabling a tetrahedron-wise linear approximation for each triangle-based distance field. The final offset surface within a tetrahedral range can be traced by slicing the tetrahedron with planes. As illustrated in the teaser figure, a key advantage of our algorithm is its ability to precisely preserve sharp features. Furthermore, this paper addresses the problem of simplifying the offset surface's complexity while preserving sharp features, formulating it as a maximal-clique problem.

Reproducing realistic collective behaviors presents a captivating yet formidable challenge. Traditional rule-based methods rely on hand-crafted principles, limiting motion diversity and realism in generated collective behaviors. Recent imitation learning methods learn from data but often require ground-truth motion trajectories and struggle with authenticity, especially in high-density groups with erratic movements. In this paper, we present a scalable approach, Collective Behavior Imitation Learning (CBIL), for learning fish schooling behavior directly from videos , without relying on captured motion trajectories. Our method first leverages Video Representation Learning, in which a Masked Video AutoEncoder (MVAE) extracts implicit states from video inputs in a self-supervised manner. The MVAE effectively maps 2D observations to implicit states that are compact and expressive for following the imitation learning stage. Then, we propose a novel adversarial imitation learning method to effectively capture complex movements of the schools of fish, enabling efficient imitation of the distribution of motion patterns measured in the latent space. It also incorporates bio-inspired rewards alongside priors to regularize and stabilize training. Once trained, CBIL can be used for various animation tasks with the learned collective motion priors. We further show its effectiveness across different species. Finally, we demonstrate the application of our system in detecting abnormal fish behavior from in-the-wild videos.

In mesh simplification, common requirements like accuracy, triangle quality, and feature alignment are often considered as a trade-off. Existing algorithms concentrate on just one or a few specific aspects of these requirements. For example, the well-known Quadric Error Metrics (QEM) approach [Garland and Heckbert 1997] prioritizes accuracy and can preserve strong feature lines/points as well, but falls short in ensuring high triangle quality and may degrade weak features that are not as distinctive as strong ones. In this paper, we propose a smooth functional that simultaneously considers all of these requirements. The functional comprises a normal anisotropy term and a Centroidal Voronoi Tessellation (CVT) [Du et al. 1999] energy term, with the variables being a set of movable points lying on the surface. The former inherits the spirit of QEM but operates in a continuous setting, while the latter encourages even point distribution, allowing various surface metrics. We further introduce a decaying weight to automatically balance the two terms. We selected 100 CAD models from the ABC dataset [Koch et al. 2019], along with 21 organic models, to compare the existing mesh simplification algorithms with ours. Experimental results reveal an important observation: the introduction of a decaying weight effectively reduces the conflict between the two terms and enables the alignment of weak features. This distinctive feature sets our approach apart from most existing mesh simplification methods and demonstrates significant potential in shape understanding. Please refer to the teaser figure for illustration.

Neural implicit representations have revolutionized dense multi-view surface reconstruction, yet their performance significantly diminishes with sparse input views. A few pioneering works have sought to tackle this challenge by leveraging additional geometric priors or multi-scene generalizability. However, they are still hindered by the imperfect choice of input views, using images under empirically determined viewpoints. We propose PVP-Recon , a novel and effective sparse-view surface reconstruction method that progressively plans the next best views to form an optimal set of sparse viewpoints for image capturing. PVP-Recon starts initial surface reconstruction with as few as 3 views and progressively adds new views which are determined based on a novel warping score that reflects the information gain of each newly added view. This progressive view planning progress is interleaved with a neural SDF-based reconstruction module that utilizes multi-resolution hash features, enhanced by a progressive training scheme and a directional Hessian loss. Quantitative and qualitative experiments on three benchmark datasets show that our system achieves high-quality reconstruction with a constrained input budget and outperforms existing baselines.

Recent advancements in 2D diffusion models allow appearance generation on untextured raw meshes. These methods create RGB textures by distilling a 2D diffusion model, which often contains unwanted baked-in shading effects and results in unrealistic rendering effects in the downstream applications. Generating Physically Based Rendering (PBR) materials instead of just RGB textures would be a promising solution. However, directly distilling the PBR material parameters from 2D diffusion models still suffers from incorrect material decomposition, such as baked-in shading effects in albedo. We introduce DreamMat , an innovative approach to resolve the aforementioned problem, to generate high-quality PBR materials from text descriptions. We find out that the main reason for the incorrect material distillation is that large-scale 2D diffusion models are only trained to generate final shading colors, resulting in insufficient constraints on material decomposition during distillation. To tackle this problem, we first finetune a new light-aware 2D diffusion model to condition on a given lighting environment and generate the shading results on this specific lighting condition. Then, by applying the same environment lights in the material distillation, DreamMat can generate high-quality PBR materials that are not only consistent with the given geometry but also free from any baked-in shading effects in albedo. Extensive experiments demonstrate that the materials produced through our methods exhibit greater visual appeal to users and achieve significantly superior rendering quality compared to baseline methods, which are preferable for downstream tasks such as game and film production.

Geometric deep learning has sparked a rising interest in computer graphics to perform shape understanding tasks, such as shape classification and semantic segmentation. When the input is a polygonal surface, one has to suffer from the irregular mesh structure. Motivated by the geometric spectral theory, we introduce Laplacian2Mesh, a novel and flexible convolutional neural network (CNN) framework for coping with irregular triangle meshes (vertices may have any valence). By mapping the input mesh surface to the multi-dimensional Laplacian-Beltrami space, Laplacian2Mesh enables one to perform shape analysis tasks directly using the mature CNNs, without the need to deal with the irregular connectivity of the mesh structure. We further define a mesh pooling operation such that the receptive field of the network can be expanded while retaining the original vertex set as well as the connections between them. Besides, we introduce a channel-wise self-attention block to learn the individual importance of feature ingredients. Laplacian2Mesh not only decouples the geometry from the irregular connectivity of the mesh structure but also better captures the global features that are central to shape classification and segmentation. Extensive tests on various datasets demonstrate the effectiveness and efficiency of Laplacian2Mesh, particularly in terms of the capability of being vulnerable to noise to fulfill various learning tasks.

Outliers will inevitably creep into the captured point cloud during 3D scanning, degrading cutting-edge models on various geometric tasks heavily. This paper looks at an intriguing question that whether point cloud completion and segmentation can promote each other to defeat outliers. To answer it, we propose a collaborative completion and segmentation network, termed CS-Net, for partial point clouds with outliers. Unlike most of existing methods, CS-Net does not need any clean (or say outlier-free) point cloud as input or any outlier removal operation. CS-Net is a new learning paradigm that makes completion and segmentation networks work collaboratively. With a cascaded architecture, our method refines the prediction progressively. Specifically, after the segmentation network, a cleaner point cloud is fed into the completion network. We design a novel completion network which harnesses the labels obtained by segmentation together with farthest point sampling to purify the point cloud and leverages KNN-grouping for better generation. Benefited from segmentation, the completion module can utilize the filtered point cloud which is cleaner for completion. Meanwhile, the segmentation module is able to distinguish outliers from target objects more accurately with the help of the clean and complete shape inferred by completion. Besides the designed collaborative mechanism of CS-Net, we establish a benchmark dataset of partial point clouds with outliers. Extensive experiments show clear improvements of our CS-Net over its competitors, in terms of outlier robustness and completion accuracy.

Surface reconstruction is a challenging task when input point clouds, especially real scans, are noisy and lack normals. Observing that the Multilayer Perceptron (MLP) and the implicit moving least-square function (IMLS) provide a dual representation of the underlying surface, we introduce Neural-IMLS, a novel approach that directly learns a noise-resistant signed distance function (SDF) from unoriented raw point clouds in a self-supervised manner. In particular, IMLS regularizes MLP by providing estimated SDFs near the surface and helps enhance its ability to represent geometric details and sharp features, while MLP regularizes IMLS by providing estimated normals. We prove that at convergence, our neural network produces a faithful SDF whose zero-level set approximates the underlying surface due to the mutual learning mechanism between the MLP and the IMLS. Extensive experiments on various benchmarks, including synthetic and real scans, show that Neural-IMLS can reconstruct faithful shapes even with noise and missing parts. The source code can be found at https://github.com/bearprin/Neural-IMLS.

Automatic tooth alignment target prediction is vital in shortening the planning time of orthodontic treatments and aligner designs. Generally, the quality of alignment targets greatly depends on the experience and ability of dentists and has enormous subjective factors. Therefore, many knowledge-driven alignment prediction methods have been proposed to help inexperienced dentists. Unfortunately, existing methods tend to directly regress tooth motion, which lacks clinical interpretability. Tooth anatomical landmarks play a critical role in orthodontics because they are effective in aiding the assessment of whether teeth are in close arrangement and normal occlusion. Thus, we consider anatomical landmark constraints to improve tooth alignment results. In this article, we present a novel tooth alignment neural network for alignment target predictions based on tooth landmark constraints and a hierarchical graph structure. We detect the landmarks of each tooth first and then construct a hierarchical graph of jaw-tooth-landmark to characterize the relationship between teeth and landmarks. Then, we define the landmark constraints to guide the network to learn the normal occlusion and predict the rigid transformation of each tooth during alignment. Our method achieves better results with the architecture built for tooth data and landmark constraints and has better explainability than previous methods with regard to clinical tooth alignments.

Volumetric data abounds in medical imaging and other fields. With the improved imaging quality and the increased resolution, volumetric datasets are getting so large that the existing tools have become inadequate for processing and analyzing the data. Here we consider the problem of computing tetrahedral meshes to represent large volumetric datasets with labeled multiple materials, which are often encountered in medical imaging or microscopy optical slice tomography. Such tetrahedral meshes are a more compact and expressive geometric representation so are in demand for efficient visualization and simulation of the data, which are impossible if the original large volumetric data are used directly due to the large memory requirement. Existing methods for meshing volumetric data are not scalable for handling large datasets due to their sheer demand on excessively large run-time memory or failure to produce a tet-mesh that preserves the multi-material structure of the original volumetric data. In this article we propose a novel approach, called Marching Windows, that uses a moving window and a disk-swap strategy to reduce the run-time memory footprint, devise a new scheme that guarantees to preserve the topological structure of the original dataset, and adopt an error-guided optimization technique to improve both geometric approximation error and mesh quality. Extensive experiments show that our method is capable of processing very large volumetric datasets beyond the capability of the existing methods and producing tetrahedral meshes of high quality.

Adversarial attack on point clouds plays a vital role in evaluating and improving the adversarial robustness of 3D deep learning models. Current attack methods are mainly applied by point perturbation in a non-manifold manner. In this paper, we formulate a novel manifold attack, which deforms the underlying 2-manifold surfaces via parameter plane stretching to generate adversarial point clouds. First, we represent the mapping between the parameter plane and underlying surface using generative-based networks. Second, the stretching is learned in the 2D parameter domain such that the generated 3D point cloud fools a pretrained classifier with minimal geometric distortion. Extensive experiments show that adversarial point clouds generated by manifold attack are smooth, undefendable and transferable, and outperform those samples generated by the state-of-the-art non-manifold ones.

Guanhua Chen, Lu Hou, Yun Chen, Wenliang Dai, Lifeng Shang, Xin Jiang, Qun Liu, Jia Pan, Wenping Wang. Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2023.

Understanding dynamic hand motions and actions from egocentric RGB videos is a fundamental yet challenging task due to self-occlusion and ambiguity. To address occlusion and ambiguity, we develop a transformer-based framework to exploit temporal information for robust estimation. Noticing the different temporal granularity of and the semantic correlation between hand pose estimation and action recognition, we build a network hierarchy with two cascaded transformer encoders, where the first one exploits the short-term temporal cue for hand pose estimation, and the latter aggregates per-frame pose and object information over a longer time span to recognize the action. Our approach achieves competitive results on two first-person hand action benchmarks, namely FPHA and H2O. Extensive ablation studies verify our design choices.

This paper presents a novel grid-based NeRF called F2- NeRF (Fast-Free-NeRF) for novel view synthesis, which enables arbitrary input camera trajectories and only costs a few minutes for training. Existing fast grid-based NeRF training frameworks, like Instant-NGP, Plenoxels, DVGO, or TensoRF, are mainly designed for bounded scenes and rely on space warping to handle unbounded scenes. Existing two widely-used space-warping methods are only designed for the forward-facing trajectory or the 360° object-centric trajectory but cannot process arbitrary trajectories. In this paper, we delve deep into the mechanism of space warping to handle unbounded scenes. Based on our analysis, we further propose a novel space-warping method called perspective warping, which allows us to handle arbitrary trajectories in the grid-based NeRF framework. Extensive experiments demonstrate that F2-NeRF is able to use the same perspective warping to render high-quality images on two standard datasets and a new free trajectory dataset collected by us. Project page: totoro97.github.io/projects/f2-nerf.

Contrastive Language-Image Pre-training (CLIP) achieves promising results in 2D zero-shot and few-shot learning. Despite the impressive performance in 2D, applying CLIP to help the learning in 3D scene understanding has yet to be explored. In this paper, we make the first attempt to investigate how CLIP knowledge benefits 3D scene understanding. We propose CLIP2Scene, a simple yet effective framework that transfers CLIP knowledge from 2D image-text pre-trained models to a 3D point cloud network. We show that the pre-trained 3D network yields impressive performance on various downstream tasks, i.e., annotation-free and fine-tuning with labelled data for semantic segmentation. Specifically, built upon CLIP, we design a Semantic-driven Cross-modal Contrastive Learning framework that pre-trains a 3D network via semantic and spatial-temporal consistency regularization. For the former, we first leverage CLIP's text semantics to select the positive and negative point samples and then employ the contrastive loss to train the 3D network. In terms of the latter, we force the consistency between the temporally coherent point cloud features and their corresponding image features. We conduct experiments on SemanticKITTI, nuScenes, and ScanNet. For the first time, our pre-trained network achieves annotation-free 3D semantic segmentation with 20.8% and 25.08% mIoU on nuScenes and ScanNet, respectively. When fine-tuned with 1% or 100% labelled data, our method significantly outperforms other self-supervised methods, with improvements of 8% and 1% mIoU, respectively. Furthermore, we demonstrate the generalizability for handling cross-domain datasets. Code is publicly available11https://github.com/runnanchen/CLIP2Scene..

We present a novel method, called NeuralUDF, for reconstructing surfaces with arbitrary topologies from 2D images via volume rendering. Recent advances in neural rendering based reconstruction have achieved compelling results. However, these methods are limited to objects with closed surfaces since they adopt Signed Distance Function (SDF) as surface representation which requires the target shape to be divided into inside and outside. In this paper, we propose to represent surfaces as the Unsigned Distance Function (UDF) and develop a new volume rendering scheme to learn the neural UDF representation. Specifically, a new density function that correlates the property of UDF with the volume rendering scheme is introduced for robust optimization of the UDF fields. Experiments on the DTU and DeepFashion3D datasets show that our method not only enables high-quality reconstruction of non-closed shapes with complex typologies, but also achieves comparable performance to the SDF based methods on the reconstruction of closed surfaces. Visit our project page at https://www.xxlong.site/NeuralUDF.

In this paper, we present a new method for the multi-view registration of point cloud. Previous multiview registration methods rely on exhaustive pairwise registration to construct a densely-connected pose graph and apply Iteratively Reweighted Least Square (IRLS) on the pose graph to compute the scan poses. However, constructing a densely-connected graph is time-consuming and contains lots of outlier edges, which makes the subsequent IRLS struggle to find correct poses. To address the above problems, we first propose to use a neural network to estimate the overlap between scan pairs, which enables us to construct a sparse but reliable pose graph. Then, we design a novel history reweighting function in the IRLS scheme, which has strong robustness to outlier edges on the graph. In comparison with existing multiview registration methods, our method achieves 11% higher registration recall on the 3DMatch dataset and ~ 13% lower registration errors on the ScanNet dataset while reducing ~ 70% required pairwise registrations. Comprehensive ablation studies are conducted to demonstrate the effectiveness of our designs. The source code is available at https://github.com/WHU-USI3DV/SGHR.

In this paper, we introduce a set of simple yet effective TOken REduction (TORE) strategies for Transformer-based Human Mesh Recovery from monocular images. Current SOTA performance is achieved by Transformer-based structures. However, they suffer from high model complexity and computation cost caused by redundant tokens. We propose token reduction strategies based on two important aspects, i.e., the 3D geometry structure and 2D image feature, where we hierarchically recover the mesh geometry with priors from body structure and conduct token clustering to pass fewer but more discriminative image feature tokens to the Transformer. Our method massively reduces the number of tokens involved in high-complexity interactions in the Transformer. This leads to a significantly reduced computational cost while still achieving competitive or even higher accuracy in shape recovery. Extensive experiments across a wide range of benchmarks validate the superior effectiveness of the proposed method. We further demonstrate the generalizability of our method on hand mesh recovery. Visit our project page at https://frank-zy-dou.github.io/projects/Tore/index.html.

We present a novel method called NeTO, for capturing the 3D geometry of solid transparent objects from 2D images via volume rendering. Reconstructing transparent objects is a very challenging task, which is ill-suited for general-purpose reconstruction techniques due to the specular light transport phenomena. Although existing refraction-tracing-based methods, designed especially for this task, achieve impressive results, they still suffer from unstable optimization and loss of fine details since the explicit surface representation they adopted is difficult to be optimized, and the self-occlusion problem is ignored for refraction-tracing. In this paper, we propose to leverage implicit Signed Distance Function (SDF) as surface representation and optimize the SDF field via volume rendering with a self-occlusion aware refractive ray tracing. The implicit representation enables our method to be capable of reconstructing high-quality reconstruction even with a limited set of views, and the self-occlusion aware strategy makes it possible for our method to accurately reconstruct the self-occluded regions. Experiments show that our method achieves faithful reconstruction results and outperforms prior works by a large margin. Visit our project page at https://www.xxlong.site/NeTO/.

We present a learning-based method, namely GeoUDF, to tackle the long-standing and challenging problem of reconstructing a discrete surface from a sparse point cloud. To be specific, we propose a geometry-guided learning method for UDF and its gradient estimation that explicitly formulates the unsigned distance of a query point as the learnable affine averaging of its distances to the tangent planes of neighboring points on the surface. Besides, we model the local geometric structure of the input point clouds by explicitly learning a quadratic polynomial for each point. This not only facilitates upsampling the input sparse point cloud but also naturally induces unoriented normal, which further augments UDF estimation. Finally, to extract triangle meshes from the predicted UDF we propose a customized edge-based marching cube module. We conduct extensive experiments and ablation studies to demonstrate the significant advantages of our method over state-of-the-art methods in terms of reconstruction accuracy, efficiency, and generality. The source code is publicly available at https://github.com/rsy6318/GeoUDF.

3D reconstruction techniques have widely been used for digital documentation of archaeological fragments. However, efficient digital capture of fragments remains as a challenge. In this work, we aim to develop a portable, high-throughput, and accurate reconstruction system for efficient digitization of fragments excavated in archaeological sites. To realize high-throughput digitization of large numbers of objects, an effective strategy is to perform scanning and reconstruction in batches. However, effective batch-based scanning and reconstruction face two key challenges: 1) how to correlate partial scans of the same object from multiple batch scans, and 2) how to register and reconstruct complete models from partial scans that exhibit only small overlaps. To tackle these two challenges, we develop a new batch-based matching algorithm that pairs the front and back sides of the fragments, and a new Bilateral Boundary ICP algorithm that can register partial scans sharing very narrow overlapping regions. Extensive validation in labs and testing in excavation sites demonstrate that these designs enable efficient batch-based scanning for fragments. We show that such a batch-based scanning and reconstruction pipeline can have immediate applications on digitizing sherds in archaeological excavations. Our project page: https://jiepengwang.github.io/FIRES/.

We propose a method, named DualMesh-UDF, to extract a surface from unsigned distance functions (UDFs), encoded by neural networks, or neural UDFs. Neural UDFs are becoming increasingly popular for surface representation because of their versatility in presenting surfaces with arbitrary topologies, as opposed to the signed distance function that is limited to representing a closed surface. However, the applications of neural UDFs are hindered by the notorious difficulty in extracting the target surfaces they represent. Recent methods for surface extraction from a neural UDF suffer from significant geometric errors or topological artifacts due to two main difficulties: (1) A UDF does not exhibit sign changes; and (2) A neural UDF typically has substantial approximation errors.DualMesh-UDF addresses these two difficulties. Specifically, given a neural UDF encoding a target surface $\bar S$ to be recovered, we first estimate the tangent planes of $\bar S$ at a set of sample points close to $\bar S$. Next, we organize these sample points into local clusters, and for each local cluster, solve a linear least squares problem to determine a final surface point. These surface points are then connected to create the output mesh surface, which approximates the target surface. The robust estimation of the tangent planes of the target surface and the subsequent minimization problem constitute our core strategy, which contributes to the favorable performance of DualMesh-UDF over other competing methods. To efficiently implement this strategy, we employ an adaptive Octree. Within this framework, we estimate the location of a surface point in each of the octree cells identified as containing part of the target surface. Extensive experiments show that our method outperforms existing methods in terms of surface reconstruction quality while maintaining comparable computational efficiency.

Deep neural networks often exhibit the overconfidence issue when encountering out-of-distribution (OOD) samples. To address this, leveraging large-scale pre-trained models like CLIP has shown promise. While CLIP has the capability to encode a vast array of interconnected concepts, current OOD detection methods based on it primarily focus on ID categories and a limited set of OOD categories. In this paper, we propose a novel approach that harnesses the power of WordNet to fully exploit the rich knowledge encapsulated within CLIP, resulting in enhanced OOD detection performance. Our methodology involves constructing a word tree that includes both in-distribution (ID) words and a large set of semantically similar OOD words selected from WordNet. By matching a test image with the concepts of the words in the word tree using CLIP, we estimate the probability of the image being classified as either ID or OOD. Furthermore, we introduce a conditional random field model to effectively handle both the parent-child and the sibling-sibling conflicts in the concept matching results. Extensive experiments under various ID/OOD settings demonstrate the effectiveness of our approach and its superiority over state-of-the-art methods.

Deep neural networks (DNNs) are dominating various computer vision solutions. However, DNN classifiers suffer from the out-of-distribution (OOD) overconfidence issue, i.e., making overconfident predictions on OOD samples. In this paper, we consider a new OOD attack task, i.e., generating OOD examples that fool DNN classifiers to trap into this issue. Specifically, we first generate seed examples by sampling from common OOD distributions, and then lift the prediction to be overconfident. Extensive experiments with different seeds and confidence-lifting solutions under white-and black-box settings validate the feasibility of OOD attack. Besides, we demonstrate its usefulness in evaluating OOD detection and alleviating the OOD overconfidence issue.

We investigate transductive zero-shot point cloud semantic segmentation, where the network is trained on seen objects and able to segment unseen objects. The 3D geometric elements are essential cues to imply a novel 3D object type. However, previous methods neglect the fine-grained relationship between the language and the 3D geometric elements. To this end, we propose a novel framework to learn the geometric primitives shared in seen and unseen categories' objects and employ a fine-grained alignment between language and the learned geometric primitives. Therefore, guided by language, the network recognizes the novel objects represented with geometric primitives. Specifically, we formulate a novel point visual representation, the similarity vector of the point's feature to the learnable prototypes, where the prototypes automatically encode geometric primitives via back-propagation. Besides, we propose a novel Unknown-aware InfoNCE Loss to fine-grained align the visual representation with language. Extensive experiments show that our method significantly outperforms other state-of-the-art methods in the harmonic mean-intersection-over-union (hIoU), with the improvement of 17.8%, 30.4%, 9.2% and 7.9% on S3DIS, ScanNet, SemanticKITTI and nuScenes datasets, respectively. Codes are available1 https://github.com/runnanchen/Zero-Shot-Point-Cloud-Segmentation.

Vision foundation models such as Contrastive Vision-Language Pre-training (CLIP) and Segment Anything (SAM) have demonstrated impressive zero-shot performance on image classification and segmentation tasks. However, the incorporation of CLIP and SAM for label-free scene understanding has yet to be explored. In this paper, we investigate the potential of vision foundation models in enabling networks to comprehend 2D and 3D worlds without labelled data. The primary challenge lies in effectively supervising networks under extremely noisy pseudo labels, which are generated by CLIP and further exacerbated during the propagation from the 2D to the 3D domain. To tackle these challenges, we propose a novel Cross-modality Noisy Supervision (CNS) method that leverages the strengths of CLIP and SAM to supervise 2D and 3D networks simultaneously. In particular, we introduce a prediction consistency regularization to co-train 2D and 3D networks, then further impose the networks' latent space consistency using the SAM's robust feature representation. Experiments conducted on diverse indoor and outdoor datasets demonstrate the superior performance of our method in understanding 2D and 3D open environments. Our 2D and 3D network achieves label-free semantic segmentation with 28.4\% and 33.5\% mIoU on ScanNet, improving 4.7\% and 7.9\%, respectively. For nuImages and nuScenes datasets, the performance is 22.1\% and 26.8\% with improvements of 3.5\% and 6.0\%, respectively. Code is available. (https://github.com/runnanchen/Label-Free-Scene-Understanding)

The Signed Distance Function (SDF), as an implicit surface representation, provides a crucial method for reconstructing a watertight surface from unorganized point clouds. The SDF has a fundamental relationship with the principles of surface vector calculus. Given a smooth surface, there exists a thin-shell space in which the SDF is differentiable everywhere such that the gradient of the SDF is an eigenvector of its Hessian matrix, with a corresponding eigenvalue of zero. In this paper, we introduce a method to directly learn the SDF from point clouds in the absence of normals. Our motivation is grounded in a fundamental observation: aligning the gradient and the Hessian of the SDF provides a more efficient mechanism to govern gradient directions. This, in turn, ensures that gradient changes more accurately reflect the true underlying variations in shape. Extensive experimental results demonstrate its ability to accurately recover the underlying shape while effectively suppressing the presence of ghost geometry.

Geodesics play a critical role in many geometry processing applications. Traditional algorithms for computing geodesics on 3D mesh models are often inefficient and slow, which make them impractical for scenarios requiring extensive querying of arbitrary point-to-point geodesics. Recently, deep implicit functions have gained popularity for 3D geometry representation, yet there is still no research on neural implicit representation of geodesics. To bridge this gap, we make the first attempt to represent geodesics using implicit learning frameworks. Specifically, we propose neural geodesic field (NeuroGF), which can be learned to encode all-pairs geodesics of a given 3D mesh model, enabling to efficiently and accurately answer queries of arbitrary point-to-point geodesic distances and paths. Evaluations on common 3D object models and real-captured scene-level meshes demonstrate our exceptional performances in terms of representation accuracy and querying efficiency. Besides, NeuroGF also provides a convenient way of jointly encoding both 3D geometry and geodesics in a unified representation. Moreover, the working mode of per-model overfitting is further extended to generalizable learning frameworks that can work on various input formats such as unstructured point clouds, which also show satisfactory performances for unseen shapes and categories. Our code and data are available at https://github.com/keeganhk/NeuroGF.

We present C · ASE, an efficient and effective framework that learns Conditional Adversarial Skill Embeddings for physics-based characters. C · ASE enables the physically simulated character to learn a diverse repertoire of skills while providing controllability in the form of direct manipulation of the skills to be performed. This is achieved by dividing the heterogeneous skill motions into distinct subsets containing homogeneous samples for training a low-level conditional model to learn the conditional behavior distribution. The skill-conditioned imitation learning naturally offers explicit control over the character’s skills after training. The training course incorporates the focal skill sampling, skeletal residual forces, and element-wise feature masking to balance diverse skills of varying complexities, mitigate dynamics mismatch to master agile motions and capture more general behavior characteristics, respectively. Once trained, the conditional model can produce highly diverse and realistic skills, outperforming state-of-the-art models, and can be repurposed in various downstream tasks. In particular, the explicit skill control handle allows a high-level policy or a user to direct the character with desired skill specifications, which we demonstrate is advantageous for interactive character animation.

Shape manipulation is a central research topic in computer graphics. Topology editing, such as breaking apart connections, joining disconnected ends, and filling/opening a topological hole, is generally more challenging than geometry editing. In this paper, we observe that the saddle points of the signed distance function (SDF) provide useful hints for altering surface topology deliberately. Based on this key observation, we parameterize the SDF into a cubic trivariate tensor-product B-spline function F whose saddle points {si} can be quickly exhausted based on a subdivision-based root-finding technique coupled with Newton’s method. Users can select one of the candidate points, say si, to edit the topology in real time. In implementation, we add a compactly supported B-spline function rooted at si, which we call a deformer in this paper, to F, with its local coordinate system aligning with the three eigenvectors of the Hessian. Combined with ray marching technique, our interactive system operates at 30 FPS. Additionally, our system empowers users to create desired bulges or concavities on the surface. An extensive user study indicates that our system is user-friendly and intuitive to operate. We demonstrate the effectiveness and usefulness of our system in a range of applications, including fixing surface reconstruction errors, artistic work design, 3D medical imaging and simulation, and antiquity restoration. Please refer to the attached video for a demonstration.

We present RoReg, a novel point cloud registration framework that fully exploits oriented descriptors and estimated local rotations in the whole registration pipeline. Previous methods mainly focus on extracting rotation-invariant descriptors for registration but unanimously neglect the orientations of descriptors. In this paper, we show that the oriented descriptors and the estimated local rotations are very useful in the whole registration pipeline, including feature description, feature detection, feature matching, and transformation estimation. Consequently, we design a novel oriented descriptor RoReg-Desc and apply RoReg-Desc to estimate the local rotations. Such estimated local rotations enable us to develop a rotation-guided detector, a rotation coherence matcher, and a one-shot-estimation RANSAC, all of which greatly improve the registration performance. Extensive experiments demonstrate that RoReg achieves state-of-the-art performance on the widely-used 3DMatch and 3DLoMatch datasets, and also generalizes well to the outdoor ETH dataset. In particular, we also provide in-depth analysis on each component of RoReg, validating the improvements brought by oriented descriptors and the estimated local rotations. Source code and supplementary material are available at https://github.com/HpWang-whu/RoReg.

Blind face restoration aims at recovering high-quality face images from those with unknown degradations. Current algorithms mainly introduce priors to complement high-quality details and achieve impressive progress. However, most of these algorithms ignore abundant contextual information in the face and its interplay with the priors, leading to sub-optimal performance. Moreover, they pay less attention to the gap between the synthetic and real-world scenarios, limiting the robustness and generalization to real-world applications. In this work, we propose RestoreFormer++, which on the one hand introduces fully-spatial attention mechanisms to model the contextual information and the interplay with the priors, and on the other hand, explores an extending degrading model to help generate more realistic degraded face images to alleviate the synthetic-to-real-world gap. Compared with current algorithms, RestoreFormer++ has several crucial benefits. First, instead of using a multi-head self-attention mechanism like the traditional visual transformer, we introduce multi-head cross-attention over multi-scale features to fully explore spatial interactions between corrupted information and high-quality priors. In this way, it can facilitate RestoreFormer++ to restore face images with higher realness and fidelity. Second, in contrast to the recognition-oriented dictionary, we learn a reconstruction-oriented dictionary as priors, which contains more diverse high-quality facial details and better accords with the restoration target. Third, we introduce an extending degrading model that contains more realistic degraded scenarios for training data synthesizing, and thus helps to enhance the robustness and generalization of our RestoreFormer++ model. Extensive experiments show that RestoreFormer++ outperforms state-of-the-art algorithms on both synthetic and real-world datasets.

We present a neural rendering-based method called NeRO for reconstructing the geometry and the BRDF of reflective objects from multiview images captured in an unknown environment. Multiview reconstruction of reflective objects is extremely challenging because specular reflections are view-dependent and thus violate the multiview consistency, which is the cornerstone for most multiview reconstruction methods. Recent neural rendering techniques can model the interaction between environment lights and the object surfaces to fit the view-dependent reflections, thus making it possible to reconstruct reflective objects from multiview images. However, accurately modeling environment lights in the neural rendering is intractable, especially when the geometry is unknown. Most existing neural rendering methods, which can model environment lights, only consider direct lights and rely on object masks to reconstruct objects with weak specular reflections. Therefore, these methods fail to reconstruct reflective objects, especially when the object mask is not available and the object is illuminated by indirect lights. We propose a two-step approach to tackle this problem. First, by applying the split-sum approximation and the integrated directional encoding to approximate the shading effects of both direct and indirect lights, we are able to accurately reconstruct the geometry of reflective objects without any object masks. Then, with the object geometry fixed, we use more accurate sampling to recover the environment lights and the BRDF of the object. Extensive experiments demonstrate that our method is capable of accurately reconstructing the geometry and the BRDF of reflective objects from only posed RGB images without knowing the environment lights and the object masks. Codes and datasets are available at https://github.com/liuyuan-pal/NeRO.

Neural implicit representation is a promising approach for reconstructing surfaces from point clouds. Existing methods combine various regularization terms, such as the Eikonal and Laplacian energy terms, to enforce the learned neural function to possess the properties of a Signed Distance Function (SDF). However, inferring the actual topology and geometry of the underlying surface from poor-quality unoriented point clouds remains challenging. In accordance with Differential Geometry, the Hessian of the SDF is singular for points within the differential thin-shell space surrounding the surface. Our approach enforces the Hessian of the neural implicit function to have a zero determinant for points near the surface. This technique aligns the gradients for a near-surface point and its on-surface projection point, producing a rough but faithful shape within just a few iterations. By annealing the weight of the singular-Hessian term, our approach ultimately produces a high-fidelity reconstruction result. Extensive experimental results demonstrate that our approach effectively suppresses ghost geometry and recovers details from unoriented point clouds with better expressiveness than existing fitting-based methods.

Estimating normals with globally consistent orientations for a raw point cloud has many downstream geometry processing applications. Despite tremendous efforts in the past decades, it remains challenging to deal with an unoriented point cloud with various imperfections, particularly in the presence of data sparsity coupled with nearby gaps or thin-walled structures. In this paper, we propose a smooth objective function to characterize the requirements of an acceptable winding-number field, which allows one to find the globally consistent normal orientations starting from a set of completely random normals. By taking the vertices of the Voronoi diagram of the point cloud as examination points, we consider the following three requirements: (1) the winding number is either 0 or 1, (2) the occurrences of 1 and the occurrences of 0 are balanced around the point cloud, and (3) the normals align with the outside Voronoi poles as much as possible. Extensive experimental results show that our method outperforms the existing approaches, especially in handling sparse and noisy point clouds, as well as shapes with complex geometry/topology.

Most of the existing point-to-mesh distance query solvers, such as Proximity Query Package (PQP), Embree and Fast Closest Point Query (FCPW), are based on bounding volume hierarchy (BVH). The hierarchical organizational structure enables one to eliminate the vast majority of triangles that do not help find the closest point. In this paper, we develop a totally different algorithmic paradigm, named P2M , to speed up point-to-mesh distance queries. Our original intention is to precompute a KD tree (KDT) of mesh vertices to approximately encode the geometry of a mesh surface containing vertices, edges and faces. However, it is very likely that the closest primitive to the query point is an edge e (resp., a face f ), but the KDT reports a mesh vertex υ instead. We call υ an interceptor of e (resp., f ). The main contribution of this paper is to invent a simple yet effective interception inspection rule and an efficient flooding interception inspection algorithm for quickly finding out all the interception pairs. Once the KDT and the interception table are precomputed, the query stage proceeds by first searching the KDT and then looking up the interception table to retrieve the closest geometric primitive. Statistics show that our query algorithm runs many times faster than the state-of-the-art solvers.

Geodesics measure the shortest distance (either locally or globally) between two points on a curved surface and serve as a fundamental tool in digital geometry processing. Suppose that we have a parameterized path$\gamma (t)=\mathbf {x}(u(t),v(t))$on a surface$\mathbf {x}=\mathbf {x}(u,v)$with$\gamma (0)=p$and$\gamma (1)=q$. We formulate the two-point geodesic problem into a minimization problem$\int _0^1 H(\Vert \mathbf {x}_uu^{\prime }(t)+\mathbf {x}_vv^{\prime }(t)\Vert)\text{d}t$, where$H(s)$satisfies$H(0)=0,H^{\prime }(s)>0$and$H^{\prime \prime }(s)\geq 0$for$s>0$. In our implementation, we choose$H(s)=e^{s^2}-1$and show that it has several unique advantages over other choices such as$H(s)=s^2$and$H(s)=s$. It is also a minimizer of the traditional geodesic length variational and able to guarantee the uniqueness and regularity in terms of curve parameterization. In the discrete setting, we construct the initial path by a sequence of moveable points$\lbrace x_i\rbrace _{i=1}^n$and minimize$\sum _{i=1}^{n} H(\Vert x_i - x_{i+1}\Vert)$. The resulting points are evenly spaced along the path. It’s obvious that our algorithm can deal with parametric surfaces. Considering that meshes, point clouds and implicit surfaces can be transformed into a signed distance function (SDF), we also discuss its implementation on a general SDF. Finally, we show that our method can be extended to solve a general least-cost path problem. We validate the proposed algorithm in terms of accuracy, performance and scalability, and demonstrate the advantages by extensive comparisons.

Creating 3D shapes from 2D drawings is an important problem with applications in content creation for computer animation and virtual reality. We introduce a new sketch-based system, CreatureShop, that enables amateurs to create high-quality textured 3D character models from 2D drawings with ease and efficiency. CreatureShop takes an input bitmap drawing of a character (such as an animal or other creature), depicted from an arbitrary descriptive pose and viewpoint, and creates a 3D shape with plausible geometric details and textures from a small number of user annotations on the 2D drawing. Our key contributions are a novel oblique view modeling method, a set of systematic approaches for producing plausible textures on the invisible or occluded parts of the 3D character (as viewed from the direction of the input drawing), and a user-friendly interactive system. We validate our system and methods by creating numerous 3D characters from various drawings, and compare our results with related works to show the advantages of our method. We perform a user study to evaluate the usability of our system, which demonstrates that our system is a practical and efficient approach to create fully-textured 3D character models for novice users.

Guanhua Chen, Shuming Ma, Yun Chen, Dongdong Zhang, Jia Pan, Wenping Wang, Furu Wei. Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2022.

We propose a method for self-supervised image representation learning under the guidance of 3D geometric consistency. Our intuition is that 3D geometric consistency priors such as smooth regions and surface discontinuities may imply consistent semantics or object boundaries, and can act as strong cues to guide the learning of 2D image representations without semantic labels. Specifically, we introduce 3D geometric consistency into a contrastive learning framework to enforce the feature consistency within image views. We propose to use geometric consistency sets as constraints and adapt the InfoNCE loss accordingly. We show that our learned image representations are general. By fine-tuning our pre-trained representations for various 2D image-based downstream tasks, including semantic segmentation, object detection, and instance segmentation on real-world indoor scene datasets, we achieve superior performance compared with state-of-the-art methods.

Speech-driven 3D facial animation is challenging due to the complex geometry of human faces and the limited availability of 3D audio-visual data. Prior works typically focus on learning phoneme-level features of short audio windows with limited context, occasionally resulting in inaccurate lip movements. To tackle this limitation, we propose a Transformer-based autoregressive model, Face-Former, which encodes the long-term audio context and autoregressively predicts a sequence of animated 3D face meshes. To cope with the data scarcity issue, we integrate the self-supervised pre-trained speech representations. Also, we devise two biased attention mechanisms well suited to this specific task, including the biased cross-modal multi-head (MH) attention and the biased causal MH self-attention with a periodic positional encoding strategy. The former effectively aligns the audio-motion modalities, whereas the latter offers abilities to generalize to longer audio sequences. Extensive experiments and a perceptual user study show that our approach outperforms the existing state-of-the-arts. The code and the video are available at: https://evelynfan.github.io/audio2face/

We present a new neural representation, called Neural Ray (NeuRay), for the novel view synthesis task. Recent works construct radiance fields from image features of input views to render novel view images, which enables the generalization to new scenes. However, due to occlusions, a 3D point may be invisible to some input views. On such a 3D point, these generalization methods will include inconsistent image features from invisible views, which interfere with the radiance field construction. To solve this problem, we predict the visibility of 3D points to input views within our NeuRay representation. This visibility enables the radiance field construction to focus on visible image features, which significantly improves its rendering quality. Meanwhile, a novel consistency loss is proposed to refine the visibility in NeuRay when finetuning on a specific scene. Experiments demonstrate that our approach achieves state-of-the-art performance on the novel view synthesis task when generalizing to unseen scenes and outperforms perscene optimization methods after finetuning. Project page:https://liuyuan-pal.github.io/NeuRay/

Blind face restoration is to recover a high-quality face image from unknown degradations. As face image contains abundant contextual information, we propose a method, RestoreFormer, which explores fully-spatial attentions to model contextual information and surpasses existing works that use local operators. RestoreFormer has several benefits compared to prior arts. First, unlike the conventional multi-head self-attention in previous Vision Transformers (ViTs), RestoreFormer incorporates a multi-head cross-attention layer to learn fully-spatial interactions between corrupted queries and high-quality key-value pairs. Second, the key-value pairs in ResotreFormer are sampled from a reconstruction-oriented high-quality dictionary, whose elements are rich in high-quality facial features specifically aimed for face reconstruction, leading to superior restoration results. Third, RestoreFormer outperforms advanced state-of-the-art methods on one synthetic dataset and three real-world datasets, as well as produces images with better visual quality. Code is available at https://github.com/wzhouxiff/RestoreFormer.git.

We propose a deep visuo-tactile model for real-time estimation of the liquid inside a deformable container in a proprioceptive way. We fuse two sensory modalities, i.e., the raw visual inputs from the RGB camera and the tactile cues from our specific tactile sensor without any extra sensor calibrations. The robotic system is well controlled and adjusted based on the estimation model in real time. The main contributions and novelties of our work are listed as follows: 1) Explore a proprioceptive way for liquid volume estimation by developing an end-to-end predictive model with multi-modal convolutional networks, which achieve a high precision with an error of ~ 2 ml in the experimental validation. 2) Propose a multi-task learning architecture which comprehensively considers the losses from both classification and regression tasks, and comparatively evaluate the performance of each variant on the collected data and actual robotic platform. 3) Utilize the proprioceptive robotic system to accurately serve and control the requested volume of liquid, which is continuously flowing into a deformable container in real time. 4) Adaptively adjust the grasping plan to achieve more stable grasping and manipulation according to the real-time liquid volume prediction.

In this paper, we propose a novel local descriptor-based framework, called You Only Hypothesize Once (YOHO), for the registration of two unaligned point clouds. In contrast to most existing local descriptors which rely on a fragile local reference frame to gain rotation invariance, the proposed descriptor achieves the rotation invariance by recent technologies of group equivariant feature learning, which brings more robustness to point density and noise. Meanwhile, the descriptor in YOHO also has a rotation-equivariant part, which enables us to estimate the registration from just one correspondence hypothesis. Such property reduces the searching space for feasible transformations, thus greatly improving both the accuracy and the efficiency of YOHO. Extensive experiments show that YOHO achieves superior performances with much fewer needed RANSAC iterations on four widely-used datasets, the 3DMatch/3DLoMatch datasets, the ETH dataset and the WHU-TLS dataset. More details are shown in our project page: https://hpwang-whu.github.io/YOHO/.

Abstract In this paper, we present a simple yet effective formulation called Coverage Axis for 3D shape skeletonization. Inspired by the set cover problem, our key idea is to cover all the surface points using as few inside medial balls as possible. This formulation inherently induces a compact and expressive approximation of the Medial Axis Transform (MAT) of a given shape. Different from previous methods that rely on local approximation error, our method allows a global consideration of the overall shape structure, leading to an efficient high‐level abstraction and superior robustness to noise. Another appealing aspect of our method is its capability to handle more generalized input such as point clouds and poor‐quality meshes. Extensive comparisons and evaluations demonstrate the remarkable effectiveness of our method for generating compact and expressive skeletal representation to approximate the MAT.

Visual Media were recommended by the associate editors as candidate papers for the Best Paper Award.The Editor-in-Chief then invited the three of us to serve as the committee for choosing the Best Paper.After careful discussion by the committee, the following paper is chosen as the winner of the Best Paper Award: EfficientPose: Efficient human pose estimation with neural architecture search [1] while two other papers are awarded the Honorable Mention Awards: Efficient fastest-path computations for road maps [2]Inferring object properties from human interaction and transferring them to new motions [3] The Best Paper Award Committee would like to offer congratulations to the winners, who in addition to the prestige conferred upon them by the awards, will also receive cash prizes: the Best Paper will receive US

Detecting 3D landmarks on cone-beam computed tomography (CBCT) is crucial to assessing and quantifying the anatomical abnormalities in 3D cephalometric analysis. However, the current methods are time-consuming and suffer from large biases in landmark localization, leading to unreliable diagnosis results. In this work, we propose a novel Structure-Aware Long Short-Term Memory framework (SA-LSTM) for efficient and accurate 3D landmark detection. To reduce the computational burden, SA-LSTM is designed in two stages. It first locates the coarse landmarks via heatmap regression on a down-sampled CBCT volume and then progressively refines landmarks by attentive offset regression using multi-resolution cropped patches. To boost accuracy, SA-LSTM captures global-local dependence among the cropping patches via self-attention. Specifically, a novel graph attention module implicitly encodes the landmark's global structure to rationalize the predicted position. Moreover, a novel attention-gated module recursively filters irrelevant local features and maintains high-confident local predictions for aggregating the final result. Experiments conducted on an in-house dataset and a public dataset show that our method outperforms state-of-the-art methods, achieving 1.64 mm and 2.37 mm average errors, respectively. Furthermore, our method is very efficient, taking only 0.5 seconds for inferring the whole CBCT volume of resolution 768×768×576 .

We propose a novel framework for computing the medial axis transform of 3D shapes while preserving their medial features via restricted power diagram (RPD). Medial features, including external features such as the sharp edges and corners of the input mesh surface and internal features such as the seams and junctions of medial axis, are important shape descriptors both topologically and geometrically. However, existing medial axis approximation methods fail to capture and preserve them due to the fundamentally under-sampling in the vicinity of medial features, and the difficulty to build their correct connections. In this paper we use the RPD of medial spheres and its affiliated structures to help solve these challenges. The dual structure of RPD provides the connectivity of medial spheres. The surfacic restricted power cell (RPC) of each medial sphere provides the tangential surface regions that these spheres have contact with. The connected components (CC) of surfacic RPC give us the classification of each sphere, to be on a medial sheet, a seam, or a junction. They allow us to detect insufficient sphere sampling around medial features and develop necessary conditions to preserve them. Using this RPD-based framework, we are able to construct high quality medial meshes with features preserved. Compared with existing sampling-based or voxel-based methods, our method is the first one that can preserve not only external features but also internal features of medial axes.

The task of explicit surface reconstruction is to generate a surface mesh by interpolating a given point cloud. Explicit surface reconstruction is necessary when the point cloud is required to appear exactly on the surface. However, for a non-perfect input, such as lack of normals, low density, irregular distribution, thin and tiny parts, and high genus, a robust explicit reconstruction method that can generate a high-quality manifold triangulation is missing. We propose a robust explicit surface reconstruction method that starts from an initial simple surface mesh, alternately performs a Filmsticking step and a Sculpting step of the initial mesh, and converges when the surface mesh interpolates all input points (except outliers) and remains stable. The Filmsticking is to minimize the geometric distance between the surface mesh and the point cloud through iteratively performing a restricted Voronoi diagram technique on the surface mesh, whereas the Sculpting is to bootstrap the Filmsticking iteration from local minima by applying appropriate geometric and topological changes of the surface mesh. Our algorithm is fully automatic and produces high-quality surface meshes for non-perfect inputs that are typically considered to be challenging for prior state of the art. We conducted extensive experiments on simulated scans and real scans to validate the effectiveness of our approach.

In this paper, we propose to compute Voronoi diagrams over mesh surfaces driven by an arbitrary geodesic distance solver, assuming that the input is a triangle mesh as well as a collection of sites P = { Pi } m i =1 on the surface. We propose two key techniques to solve this problem. First, as the partition is determined by minimizing the m distance fields, each of which rooted at a source site, we suggest keeping one or more distance triples, for each triangle, that may help determine the Voronoi bisectors when one uses a mark-and-sweep geodesic algorithm to predict the multi-source distance field. Second, rather than keep the distance itself at a mesh vertex, we use the squared distance to characterize the linear change of distance field restricted in a triangle, which is proved to induce an exact VD when the base surface reduces to a planar triangle mesh. Specially, our algorithm also supports the Euclidean distance, which can handle thin-sheet models (e.g. leaf) and runs faster than the traditional restricted Voronoi diagram (RVD) algorithm. It is very extensible to deal with various variants of surface-based Voronoi diagrams including (1) surface-based power diagram, (2) constrained Voronoi diagram with curve-type breaklines, and (3) curve-type generators. We conduct extensive experimental results to validate the ability to approximate the exact VD in different distance-driven scenarios.

3D Morphable models of the human body capture variations among subjects and are useful in reconstruction and editing applications. Current dental models use an explicit mesh scene representation and model only the teeth, ignoring the gum. In this work, we present the first parametric 3D morphable dental model for both teeth and gum. Our model uses an implicit scene representation and is learned from rigidly aligned scans. It is based on a component-wise representation for each tooth and the gum, together with a learnable latent code for each of such components. It also learns a template shape thus enabling several applications such as segmentation, interpolation and tooth replacement. Our reconstruction quality is on par with the most advanced global implicit representations while enabling novel applications. The code will be available at https://github.com/cong-yi/DMM

We propose a new video vectorization approach for converting videos in the raster format to vector representation with the benefits of resolution independence and compact storage. Through classifying extracted curves in each video frame into salient ones and non-salient ones, we introduce a novel bipartite diffusion curves (BDCs) representation in order to preserve both important image features such as sharp boundaries and regions with smooth color variation. This bipartite representation allows us to propagate non-salient curves across frames such that the propagation, in conjunction with geometry optimization and color optimization of salient curves, ensures the preservation of fine details within each frame and across different frames, and meanwhile, achieves good spatial-temporal coherence. Thorough experiments on a variety of videos show that our method is capable of converting videos to the vector representation with low reconstruction errors, low computational cost, and fine details, demonstrating our superior performance over the state of the art. We also show that, when used for video upsampling, our method produces results comparable to video super-resolution.

Segmenting arbitrary 3D objects into constituent parts that are structurally meaningful is a fundamental problem encountered in a wide range of computer graphics applications. Existing methods for 3D shape segmentation suffer from complex geometry processing and heavy computation caused by using low-level features and fragmented segmentation results due to the lack of global consideration. We present an efficient method, called SEG-MAT, based on the medial axis transform (MAT) of the input shape. Specifically, with the rich geometrical and structural information encoded in the MAT, we are able to develop a simple and principled approach to effectively identify the various types of junctions between different parts of a 3D shape. Extensive evaluations and comparisons show that our method outperforms the state-of-the-art methods in terms of segmentation quality and is also one order of magnitude faster.

This article presents a simple yet effective method for computing geodesic distances on triangle meshes. Unlike the popular window propagation methods that partition mesh edges into intervals of varying lengths, our method places evenly-spaced, source-independent Steiner points on edges. Given a source vertex, our method constructs a Steiner-point graph that partitions the surface into mutually exclusive tracks, called geodesic tracks. Inside each triangle, the tracks form sub-regions in which the change of distance field is approximately linear. Our method does not require any pre-computation, and can effectively balance speed and accuracy. Experimental results show that with 5 Steiner points on each edge, the mean relative error is less than 0.3 % for common 3D models used in the graphics community. We propose a set of effective filtering rules to eliminate a large amount of useless broadcast events. For a 1000K-face model, our method runs 10 times faster than the conventional Steiner point method that examines a complete graph of Steiner points in each triangle. We also observe that using more Steiner points increases the accuracy at only a small extra computational cost. Our method works well for meshes with poor triangulation and non-manifold configuration, which often poses challenges to the existing PDE methods. We show that geodesic tracks, as a new data structure that encodes rich information of discrete geodesics, support accurate geodesic path and isoline tracing, and efficient distance query. Our method can be easily extended to meshes with non-constant density functions and/or anisotropic metrics.

We introduce MulayCap, a novel human performance capture method using a monocular video camera without the need for pre-scanning. The method uses "multi-layer" representations for geometry reconstruction and texture rendering, respectively. For geometry reconstruction, we decompose the clothed human into multiple geometry layers, namely a body mesh layer and a garment piece layer. The key technique behind is a Garment-from-Video (GfV) method for optimizing the garment shape and reconstructing the dynamic cloth to fit the input video sequence, based on a cloth simulation model which is effectively solved with gradient descent. For texture rendering, we decompose each input image frame into a shading layer and an albedo layer, and propose a method for fusing a fixed albedo map and solving for detailed garment geometry using the shading layer. Compared with existing single view human performance capture systems, our "multi-layer" approach bypasses the tedious and time consuming scanning step for obtaining a human specific mesh template. Experimental results demonstrate that MulayCap produces realistic rendering of dynamically changing details that has not been achieved in any previous monocular video camera systems. Benefiting from its fully semantic modeling, MulayCap can be applied to various important editing applications, such as cloth editing, re-targeting, relighting, and AR applications.

3D sketching in virtual reality (VR) enables users to create 3D virtual objects intuitively and immersively. However, previous studies showed that mid-air drawing may lead to inaccurate sketches. To address this issue, we propose to use one hand as a canvas proxy and the index finger of the other hand as a 3D pen. To this end, we first perform a formative study to compare two-handed interaction with tablet-pen interaction for VR sketching. Based on the findings of this study, we design HandPainter, a VR sketching system which focuses on the direct use of two hands for 3D sketching without requesting any tablet, pen, or VR controller. Our implementation is based on a pair of VR gloves, which provide hand tracking and gesture capture. We devise a set of intuitive gestures to control various functionalities required during 3D sketching, such as canvas panning and drawing positioning. We show the effectiveness of HandPainter by presenting a number of sketching results and discussing the outcomes of a user study-based comparison with mid-air drawing and tablet-based sketching tools.

We introduce Point2Skeleton, an unsupervised method to learn skeletal representations from point clouds. Existing skeletonization methods are limited to tubular shapes and the stringent requirement of watertight input, while our method aims to produce more generalized skeletal representations for complex structures and handle point clouds. Our key idea is to use the insights of the medial axis transform (MAT) to capture the intrinsic geometric and topological natures of the original input points. We first predict a set of skeletal points by learning a geometric transformation, and then analyze the connectivity of the skeletal points to form skeletal mesh structures. Extensive evaluations and comparisons show our method has superior performance and robustness. The learned skeletal representation will benefit several unsupervised tasks for point clouds, such as surface reconstruction and segmentation.

Motion coherence is an important clue for distinguishing true correspondences from false ones. Modeling motion coherence on sparse putative correspondences is challenging due to their sparsity and uneven distributions. Existing works on motion coherence are sensitive to parameter settings and have difficulty in dealing with complex motion patterns. In this paper, we introduce a network called Laplacian Motion Coherence Network (LMCNet) to learn motion coherence property for correspondence pruning. We propose a novel formulation of fitting coherent motions with a smooth function on a graph of correspondences and show that this formulation allows a closed-form solution by graph Laplacian. This closed-form solution enables us to design a differentiable layer in a learning framework to capture global motion coherence from putative correspondences. The global motion coherence is further combined with local coherence extracted by another local layer to robustly detect inlier correspondences. Experiments demonstrate that LMCNet has superior performances to the state of the art in relative camera pose estimation and correspondences pruning of dynamic scenes1.

We present a novel method for multi-view depth estimation from a single video, which is a critical task in various applications, such as perception, reconstruction and robot navigation. Although previous learning-based methods have demonstrated compelling results, most works estimate depth maps of individual video frames independently, without taking into consideration the strong geometric and temporal coherence among the frames. Moreover, current state-of-the-art (SOTA) models mostly adopt a fully 3D convolution network for cost regularization and therefore require high computational cost, thus limiting their deployment in real-world applications. Our method achieves temporally coherent depth estimation results by using a novel Epipolar Spatio-Temporal (EST) transformer to explicitly associate geometric and temporal correlation with multiple estimated depth maps. Furthermore, to reduce the computational cost, inspired by recent Mixture-of-Experts models, we design a compact hybrid network consisting of a 2D context-aware network and a 3D matching network which learn 2D context information and 3D disparity cues separately. Extensive experiments demonstrate that our method achieves higher accuracy in depth estimation and significant speedup than the SOTA methods.

Previous work mainly focuses on improving cross-lingual transfer for NLU tasks with a multilingual pretrained encoder (MPE), or improving the performance on supervised machine translation with BERT.However, it is under-explored that whether the MPE can help to facilitate the cross-lingual transferability of NMT model.In this paper, we focus on a zero-shot cross-lingual transfer task in NMT.In this task, the NMT model is trained with parallel dataset of only one language pair and an off-the-shelf MPE, then it is directly tested on zero-shot language pairs.We propose SixT, a simple yet effective model for this task.SixT leverages the MPE with a two-stage training schedule and gets further improvement with a position disentangled encoder and a capacity-enhanced decoder.Using this method, SixT significantly outperforms mBART, a pretrained multilingual encoderdecoder model explicitly designed for NMT, with an average improvement of 7.1 BLEU on zero-shot any-to-English test sets across 14 source languages.Furthermore, with much less training computation cost and training data, our model achieves better performance on 15 any-to-English test sets than CRISS and m2m-100, two strong multilingual NMT baselines.

Personalized video highlight detection aims to shorten a long video to interesting moments according to a user’s preference, which has recently raised the community’s attention. Current methods regard the user’s history as holistic information to predict the user’s preference but negating the inherent diversity of the user’s interests, resulting in vague preference representation. In this paper, we propose a simple yet efficient preference reasoning framework (PR-Net) to explicitly take the diverse interests into account for frame-level highlight prediction. Specifically, distinct user-specific preferences for each input query frame are produced, presented as the similarity weighted sum of history highlights to the corresponding query frame. Next, distinct comprehensive preferences are formed by the user-specific preferences and a learnable generic preference for more overall highlight measurement. Lastly, the degree of highlight and non-highlight for each query frame is calculated as semantic similarity to its comprehensive and non-highlight preferences, respectively. Besides, to alleviate the ambiguity due to the incomplete annotation, a new bidirectional contrastive loss is proposed to ensure a compact and differentiable metric space. In this way, our method significantly outperforms state-of-the-art methods with a relative improvement of 12% in mean accuracy precision.

We present a novel method for single image depth estimation using surface normal constraints. Existing depth estimation methods either suffer from the lack of geometric constraints, or are limited to the difficulty of reliably capturing geometric context, which leads to a bottleneck of depth estimation quality. We therefore introduce a simple yet effective method, named Adaptive Surface Normal (ASN) constraint, to effectively correlate the depth estimation with geometric consistency. Our key idea is to adaptively determine the reliable local geometry from a set of randomly sampled candidates to derive surface normal constraint, for which we measure the consistency of the geometric contextual features. As a result, our method can faithfully reconstruct the 3D geometry and is robust to local shape variations, such as boundaries, sharp corners and noises. We conduct extensive evaluations and comparisons using public datasets. The experimental results demonstrate our method outperforms the state-of-the-art methods and has superior efficiency and robustness. Codes are available at: https://github.com/xxlong0/ASNDepth

Overconfident predictions on out-of-distribution (OOD) samples is a thorny issue for deep neural networks. The key to resolve the OOD overconfidence issue inherently is to build a subset of OOD samples and then suppress predictions on them. This paper proposes the Chamfer OOD examples (CODEs), whose distribution is close to that of in-distribution samples, and thus could be utilized to alleviate the OOD overconfidence issue effectively by suppressing predictions on them. To obtain CODEs, we first generate seed OOD examples via slicing&splicing operations on in-distribution samples from different categories, and then feed them to the Chamfer generative adversarial network for distribution transformation, without accessing to any extra data. Training with suppressing predictions on CODEs is validated to alleviate the OOD overconfidence issue largely without hurting classification accuracy, and outperform the state-of-the-art methods. Besides, we demonstrate CODEs are useful for improving OOD detection and classification.

This paper presents a neural network for robust normal estimation on point clouds, named AdaFit, that can deal with point clouds with noise and density variations. Existing works use a network to learn point-wise weights for weighted least squares surface fitting to estimate the normals, which has difficulty in finding accurate normals in complex regions or containing noisy points. By analyzing the step of weighted least squares surface fitting, we find that it is hard to determine the polynomial order of the fitting surface and the fitting surface is sensitive to outliers. To address these problems, we propose a simple yet effective solution that adds an additional offset prediction to improve the quality of normal estimation. Furthermore, in order to take advantage of points from different neighborhood sizes, a novel Cascaded Scale Aggregation layer is proposed to help the network predict more accurate point-wise offsets and weights. Extensive experiments demonstrate that AdaFit achieves state-of-the-art performance on both the synthetic PCPNet dataset and the real-word SceneNN dataset. The code is publicly available at https://github.com/Runsong123/AdaFit.

We study the problem of weakly supervised grounded image captioning. That is, given an image, the goal is to automatically generate a sentence describing the context of the image with each noun word grounded to the corresponding region in the image. This task is challenging due to the lack of explicit fine-grained region word alignments as supervision. Previous weakly supervised methods mainly explore various kinds of regularization schemes to improve attention accuracy. However, their performances are still far from the fully supervised ones. One main issue that has been ignored is that the attention for generating visually groundable words may only focus on the most discriminate parts and can not cover the whole object. To this end, we propose a simple yet effective method to alleviate the issue, termed as partial grounding problem in our paper. Specifically, we design a distributed attention mechanism to enforce the network to aggregate information from multiple spatially different regions with consistent semantics while generating the words. Therefore, the union of the focused region proposals should form a visual region that encloses the object of interest completely. Extensive experiments have demonstrated the superiority of our proposed method compared with the state-of-the-arts.

We present a novel neural surface reconstruction method, called NeuS, for reconstructing objects and scenes with high fidelity from 2D image inputs. Existing neural surface reconstruction approaches, such as DVR [Niemeyer et al., 2020] and IDR [Yariv et al., 2020], require foreground mask as supervision, easily get trapped in local minima, and therefore struggle with the reconstruction of objects with severe self-occlusion or thin structures. Meanwhile, recent neural methods for novel view synthesis, such as NeRF [Mildenhall et al., 2020] and its variants, use volume rendering to produce a neural scene representation with robustness of optimization, even for highly complex objects. However, extracting high-quality surfaces from this learned implicit representation is difficult because there are not sufficient surface constraints in the representation. In NeuS, we propose to represent a surface as the zero-level set of a signed distance function (SDF) and develop a new volume rendering method to train a neural SDF representation. We observe that the conventional volume rendering method causes inherent geometric errors (i.e. bias) for surface reconstruction, and therefore propose a new formulation that is free of bias in the first order of approximation, thus leading to more accurate surface reconstruction even without the mask supervision. Experiments on the DTU dataset and the BlendedMVS dataset show that NeuS outperforms the state-of-the-arts in high-quality surface reconstruction, especially for objects and scenes with complex structures and self-occlusion.

Abstract Different from other types of bas‐reliefs, a flower bas‐relief contains a large number of depth‐discontinuity edges. Most existing line‐based methods reconstruct free‐form surfaces by ignoring the depth‐discontinuities, thus are less efficient in modeling flower bas‐reliefs. This paper presents a neural‐based solution which benefits from the recent advances in CNN. Specially, we use line gradients to encode the depth orderings at leaf edges. Given a line drawing, a heuristic method is first proposed to compute 2D gradients at lines. Line gradients and dense curvatures interpolated from sparse user inputs are then fed into a neural network, which outputs depths and normals of the final bas‐relief. In addition, we introduce an object‐based method to generate flower bas‐reliefs and line drawings for network training. Extensive experiments show that our method is effective in modelling bas‐reliefs with depth‐discontinuity edges. User evaluation also shows that our method is intuitive and accessible to common users.

Superpixel segmentation, as a central image processing task, has many applications in computer vision and computer graphics. Boundary alignment and shape compactness are leading indicators to evaluate a superpixel segmentation algorithm. Furthermore, convexity can make superpixels reflect more geometric structures in images and provide a more concise over-segmentation result. In this paper, we consider generating convex and compact superpixels while satisfying the constraints of adhering to the boundary as far as possible. We formulate the new superpixel segmentation into an edge-constrained centroidal power diagram (ECCPD) optimization problem. In the implementation, we optimize the superpixel configurations by repeatedly performing two alternative operations, which include site location updating and weight updating through a weight function defined by image features. Compared with existing superpixel methods, our method can partition an image into fully convex and compact superpixels with better boundary adherence. Extensive experimental results show that our approach outperforms existing superpixel segmentation methods in boundary alignment and compactness for generating convex superpixels.

Performance degradation due to domain shift remains a major challenge in medical image analysis. Unsupervised domain adaptation that transfers knowledge learned from the source domain with ground truth labels to the target domain without any annotation is the mainstream solution to resolve this issue. In this paper, we present a novel unsupervised domain adaptation framework for cross-modality cardiac segmentation, by explicitly capturing a common cardiac structure embedded across different modalities to guide cardiac segmentation. In particular, we first extract a set of 3D landmarks, in a self-supervised manner, to represent the cardiac structure of different modalities. The high-level structure information is then combined with another complementary feature, the Canny edges, to produce accurate cardiac segmentation results both in the source and target domains. We extensively evaluate our method on the MICCAI 2017 MM-WHS dataset for cardiac segmentation. The evaluation, comparison and comprehensive ablation studies demonstrate that our approach achieves satisfactory segmentation results and outperforms state-of-the-art unsupervised domain adaptation methods by a significant margin.

We present a novel approach for completing and reconstructing 3D shapes from incomplete scanned data by using deep neural networks. Rather than being trained on supervised completion tasks and applied on a testing shape, the network is optimized from scratch on the single testing shape to fully adapt to the shape and complete the missing data using contextual guidance from the known regions. The ability to complete missing data by an untrained neural network is usually referred to as the deep prior . In this article, we interpret the deep prior from a neural tangent kernel (NTK) perspective and show that the completed shape patches by the trained CNN are naturally similar to existing patches, as they are proximate in the kernel feature space induced by NTK. The interpretation allows us to design more efficient network structures and learning mechanisms for the shape completion and reconstruction task. Being more aware of structural regularities than both traditional and other unsupervised learning-based reconstruction methods, our approach completes large missing regions with plausible shapes and complements supervised learning-based methods that use database priors by requiring no extra training dataset and showing flexible adaptation to a particular shape instance.

Motivated by the fact that the medial axis transform is able to encode the shape completely, we propose to use as few medial balls as possible to approximate the original enclosed volume by the boundary surface. We progressively select new medial balls, in a top-down style, to enlarge the region spanned by the existing medial balls. The key spirit of the selection strategy is to encourage large medial balls while imposing given geometric constraints. We further propose a speedup technique based on a provable observation that the intersection of medial balls implies the adjacency of power cells (in the sense of the power crust).We further elaborate the selection rules in combination with two closely related applications. One application is to develop an easy-to-use ball-stick modeling system that helps non-professional users to quickly build a shape with only balls and wires, but any penetration between two medial balls must be suppressed. The other application is to generate porous structures with convex, compact (with a high isoperimetric quotient) and shape-aware pores where two adjacent spherical pores may have penetration as long as the mechanical rigidity can be well preserved.

Synthesizing realistic videos of humans using neural networks has been a popular alternative to the conventional graphics-based rendering pipeline due to its high efficiency. Existing works typically formulate this as an image-to-image translation problem in 2D screen space, which leads to artifacts such as over-smoothing, missing body parts, and temporal instability of fine-scale detail, such as pose-dependent wrinkles in the clothing. In this article, we propose a novel human video synthesis method that approaches these limiting factors by explicitly disentangling the learning of time-coherent fine-scale details from the embedding of the human in 2D screen space. More specifically, our method relies on the combination of two convolutional neural networks (CNNs). Given the pose information, the first CNN predicts a dynamic texture map that contains time-coherent high-frequency details, and the second CNN conditions the generation of the final video on the temporally coherent output of the first CNN. We demonstrate several applications of our approach, such as human reenactment and novel view synthesis from monocular video, where we show significant improvement over the state of the art both qualitatively and quantitatively.

Learning structures of 3D shapes is a fundamental problem in the field of computer graphics and geometry processing. We present a simple yet interpretable unsupervised method for learning a new structural representation in the form of 3D structure points. The 3D structure points produced by our method encode the shape structure intrinsically and exhibit semantic consistency across all the shape instances with similar structures. This is a challenging goal that has not fully been achieved by other methods. Specifically, our method takes a 3D point cloud as input and encodes it as a set of local features. The local features are then passed through a novel point integration module to produce a set of 3D structure points. The chamfer distance is used as reconstruction loss to ensure the structure points lie close to the input point cloud. Extensive experiments have shown that our method outperforms the state-of-the-art on the semantic shape correspondence task and achieves comparable performance with the state-of-the-art on the segmentation label transfer task. Moreover, the PCA based shape embedding built upon consistent structure points demonstrates good performance in preserving the shape structures. Code is available at https://github.com/NolenChen/3DStructurePoints.

Facial skin texture synthesis is a fundamental problem in high-quality facial image generation and enhancement. The key behind is how to effectively synthesize plausible textured noise for the faces. With the development of CNNs and GANs, most works cast the problem as an image to image translation problem. However, these methods lack an explicit mechanism to simulate the facial noise pattern, so that the generated images are of obvious artifacts. To this end, we propose a new facial noise generation method. Specifically, we utilize the property of blue noise and Gabor filter to implicitly guide the asymmetrical sampling for the face region as a guidance map, where non-uniform point sampling is conducted. Thus we propose a novel Blue-Noise Gabor Module to produce a spatial-variant noisy image. Our proposed two-branch framework combined facial identity enhancing with textures details generation to jointly produce a high-quality facial image. Experimental results demonstrate the superiority of our method compared with the state-of-the-art, which enables the generation of high-quality facial texture based on a 2D image only, without the involvement of any 3D models.

Abstract Grayscale images are intensively used to construct or represent geometric details infield of computer graphics. In practice, displacement mapping technique often allows an 8‐bit grayscale image input to manipulate the position of vertices. Human eyes are insensitive to the change of intensity between consecutive gray levels, so a grayscale image only provides 256 levels of luminances. However, when the luminances are converted into geometric elements, certain artifacts such as false contours become obvious. In this paper, we formulate the geometric decontouring as a constrained optimization problem from a geometric perspective. Instead of directly solving this optimization problem, we propose a data‐driven method to learn a residual mapping function. We design a Geometric DeContouring Network (GDCNet) to eliminate the false contours effectively. To this end, we adopt a ResNet‐based network structure and a normal‐based loss function. Extensive experimental results demonstrate that accurate reconstructions can be achieved effectively. Our method can be used as a relief compressed representation and enhance the traditional displacement mapping technique to augment 3D models with high‐quality geometric details using grayscale images efficiently.

Abstract 3D scanned point cloud data of teeth is popular used in digital orthodontics. The classification and semantic labelling for point cloud of each tooth is a key and challenging task for planning dental treatment. Utilizing the priori ordered position information of tooth arrangement, we propose an effective network for tooth model classification in this paper. The relative position and the adjacency similarity feature vectors are calculated for tooth 3D model, and combine the geometric feature into the fully connected layers of the classification training task. For the classification of dental anomalies, we present a dental anomalies processing method to improve the classification accuracy. We also use FocalLoss as the loss function to solve the sample imbalance of wisdom teeth. The extensive evaluations, ablation studies and comparisons demonstrate that the proposed network can classify tooth models accurately and automatically and outperforms state‐of‐the‐art point cloud classification methods.

Abstract Apollonius diagrams, also known as additively weighted Voronoi diagrams, are an extension of Voronoi diagrams, where the weighted distance is defined by the Euclidean distance minus the weight. The bisectors of Apollonius diagrams have a hyperbolic form, which is fundamentally different from traditional Voronoi diagrams and power diagrams. Though robust solvers are available for computing 2D Apollonius diagrams, there is no practical approach for the 3D counterpart. In this paper, we systematically analyze the structural features of 3D Apollonius diagrams, and then develop a fast algorithm for robustly computing Apollonius diagrams in 3D. Our algorithm consists of vertex location, edge tracing and face extraction, among which the key step is to adaptively subdivide the initial large box into a set of sufficiently small boxes such that each box contains at most one Apollonius vertex. Finally, we use centroidal Voronoi tessellation (CVT) to discretize the curved bisectors with well‐tessellated triangle meshes. We validate the effectiveness and robustness of our algorithm through extensive evaluation and experiments. We also demonstrate an application on computing centroidal Apollonius diagram.

Abstract In contrast to 3D model that can be freely observed, p ortrait bas‐relief projects slightly from the background and is limited by fixed viewpoint. In this paper, we propose a novel method to reconstruct the underlying 3D shape from a single 2.5D bas‐relief, providing observers wider viewing perspectives. Our target is to make the reconstructed portrait has natural depth ordering and similar appearance to the input. To achieve this, we first use a 3D template face to fit the portrait. Then, we optimize the face shape by normal transfer and Poisson surface reconstruction. The hair and body regions are finally reconstructed and combined with the 3D face. From the resulting 3D shape, one can generate new reliefs with varying poses and thickness, freeing the input one from fixed view. A number of experimental results verify the effectiveness of our method.

Arrangements of geometric objects refer to the spatial partitions formed by the objects, and they serve as an underlining structure of motion design, analysis, and planning in CAD/CAM, robotics, molecular modeling, manufacturing, and computer-assisted radio-surgery. Arrangements are especially useful to collision detection, which is a key task in various applications such as computer animation, virtual reality, computer games, robotics, CAD/CAM, and computational physics. Ellipsoids are commonly used as bounding volumes in approximating complex geometric objects in collision detection. In this article, we present an in-depth study on the arrangements formed by two ellipsoids. Specifically, we present a classification of these arrangements and propose an efficient algorithm for determining the arrangement formed by any particular pair of ellipsoids. A stratification diagram is also established to show the connections among all the arrangements formed by two ellipsoids. Our results, for the first time, elucidate all possible relative positions between two arbitrary ellipsoids and provide an efficient and robust algorithm for determining the relative position of any two given ellipsoids, therefore providing the necessary foundation for developing practical and trustworthy methods for processing ellipsoids for collision analysis or simulation in various applications.

Thin structures, such as wire-frame sculptures, fences, cables, power lines, and tree branches, are common in the real world. It is extremely challenging to acquire their 3D digital models using traditional image-based or depth-based reconstruction methods, because thin structures often lack distinct point features and have severe self-occlusion. We propose the first approach that simultaneously estimates camera motion and reconstructs the geometry of complex 3D thin structures in high quality from a color video captured by a handheld camera. Specifically, we present a new curve-based approach to estimate accurate camera poses by establishing correspondences between featureless thin objects in the foreground in consecutive video frames, without requiring visual texture in the background scene to lock on. Enabled by this effective curve-based camera pose estimation strategy, we develop an iterative optimization method with tailored measures on geometry, topology as well as self-occlusion handling for reconstructing 3D thin structures. Extensive validations on a variety of thin structures show that our method achieves accurate camera pose estimation and faithful reconstruction of 3D thin structures with complex shape and topology at a level that has not been attained by other existing reconstruction methods.

We present a novel solution to enable portrait relief modeling from a single image. The main challenges are geometry reconstruction, facial details recovery and depth structure preservation. Previous image-based methods are developed for portrait bas-relief modeling in 2.5D form, but not adequate for 3D-like high relief modeling with undercut features. In this paper, we propose a template-based framework to generate portrait reliefs of various forms. Our method benefits from Shape-from-Shading (SFS). Specifically, we use bi-Laplacian mesh deformation to guide the relief modeling. Given a portrait image, we first use a template face to fit the portrait. We then apply bi-Laplacian mesh deformation to align the facial features. Afterwards, SFS-based reconstruction with a few user interactions is used to optimize the face depth, and create a relief with similar appearance to the input. Both depth structures and geometric details can be well constructed in the final relief. Experiments and comparisons to other methods demonstrate the effectiveness of the proposed method.

To safely and efficiently navigate in complex urban traffic, autonomous vehicles must make responsible predictions in relation to surrounding traffic-agents (vehicles, bicycles, pedestrians, etc.). A challenging and critical task is to explore the movement patterns of different traffic-agents and predict their future trajectories accurately to help the autonomous vehicle make reasonable navigation decision. To solve this problem, we propose a long short-term memory-based (LSTM-based) realtime traffic prediction algorithm, TrafficPredict. Our approach uses an instance layer to learn instances’ movements and interactions and has a category layer to learn the similarities of instances belonging to the same type to refine the prediction. In order to evaluate its performance, we collected trajectory datasets in a large city consisting of varying conditions and traffic densities. The dataset includes many challenging scenarios where vehicles, bicycles, and pedestrians move among one another. We evaluate the performance of TrafficPredict on our new dataset and highlight its higher accuracy for trajectory prediction by comparing with prior prediction methods.

This paper proposes a method that uses deep convolutional neural networks to achieve automatic and accurate tooth instance segmentation and identification from CBCT (cone beam CT) images for digital dentistry. The core of our method is a two-stage network. In the first stage, an edge map is extracted from the input CBCT image to enhance image contrast along shape boundaries. Then this edge map and the input images are passed to the second stage. In the second stage, we build our network upon the 3D region proposal network (RPN) with a novel learned-similarity matrix to help efficiently remove redundant proposals, speed up training and save GPU memory. To resolve the ambiguity in the identification task, we encode teeth spatial relationships as an additional feature input in the identification task, which helps to remarkably improve the identification accuracy. Our evaluation, comparison and comprehensive ablation studies demonstrate that our method produces accurate instance segmentation and identification results automatically and outperforms the state-of-the-art approaches. To the best of our knowledge, our method is the first to use neural networks to achieve automatic tooth segmentation and identification from CBCT images.

We present a novel approach to align partial 3D reconstructions which may not have substantial overlap. Using floorplan priors, our method jointly predicts a room layout and estimates the transformations from a set of partial 3D data. Unlike the existing methods relying on feature descriptors to establish correspondences, we exploit the 3D "box" structure of a typical room layout that meets the Manhattan World property. We first estimate a local layout for each partial scan separately and then combine these local layouts to form a globally aligned layout with loop closure. Without the requirement of feature matching, the proposed method enables some novel applications ranging from large or featureless scene reconstruction and modeling from sparse input. We validate our method quantitatively and qualitatively on real and synthetic scenes of various sizes and complexities. The evaluations and comparisons show superior effectiveness and accuracy of our method.

3D deep learning performance depends on object representation and local feature extraction. In this work, we present MAT-Net, a neural network which captures local and global features from the Medial Axis Transform (MAT). Different from K-Nearest-Neighbor method which extracts local features by a fixed number of neighbors, our MAT-Net exploits effective modules Group-MAT and Edge-Net to process topological structure. Experimental results illustrate that MAT-Net demonstrates competitive or better performance on 3D shape recognition than state-of-the-art methods, and prove that MAT representation has excellent capacity in 3D deep learning, even in the case of low resolution.

Abstract Freeform surfaces whose principal curvature line network is regularly distributed, are essential to many real applications like CAD modeling, architecture design, and industrial fabrication. However, most designed surfaces do not hold this nice property because it is hard to enforce such constraints in the design process. In this paper, we present a novel method for surface fairing which takes a regular distribution of the principal curvature line network on a surface as an objective. Our method first removes the high‐frequency signals from the curvature tensor field of an input freeform surface by a novel rolling guidance tensor filter, which results in a more regular and smooth curvature tensor field, then deforms the input surface to match the smoothed field as much as possible. As an application, we solve the problem of approximating freeform surfaces with regular principal curvature line networks, discretized by quadrilateral meshes. By introducing the circular or conical conditions on the quadrilateral mesh to guarantee the existence of discrete principal curvature line networks, and minimizing the approximate error to the original surface and improving the fairness of the quad mesh, we obtain a regular discrete principal curvature line network that approximates the original surface. We evaluate the efficacy of our method on various freeform surfaces and demonstrate the superiority of the rolling guidance tensor filter over other tensor smoothing techniques. We also utilize our method to generate high‐quality circular/conical meshes for architecture design and cyclide spline surfaces for CAD modeling.

3D mesh models created by human users and shared through online platforms and datasets flourish recently. While the creators generally have spent large efforts in modeling the visually appealing shapes with both large scale structures and intricate details, a majority of the meshes are unfortunately flawed in terms of having duplicate faces, mis-oriented regions, disconnected patches, etc., due to multiple factors involving both human errors and software inconsistencies. All these artifacts have severely limited the possible low-level and high-level processing tasks that can be applied to the rich datasets. In this work, we present a novel approach to fix these man-made meshes such that the outputs are guaranteed to be oriented manifold meshes that preserve the original structures, big and small, as much as possible. Our key observation is that the models all visually look meaningful, which leads to our strategy of repairing the flaws while always preserving the visual quality. We apply local refinements and removals only where necessary to achieve minimal intrusion of the original meshes, and global adjustments through robust optimization to ensure the outputs are valid manifold meshes with optimal connections. We test the approach on large-scale 3D datasets, and obtain quality meshes that are more readily usable for further geometry processing tasks.

We propose a method for generating video-realistic animations of real humans under user control. In contrast to conventional human character rendering, we do not require the availability of a production-quality photo-realistic three-dimensional (3D) model of the human but instead rely on a video sequence in conjunction with a (medium-quality) controllable 3D template model of the person. With that, our approach significantly reduces production cost compared to conventional rendering approaches based on production-quality 3D models and can also be used to realistically edit existing videos. Technically, this is achieved by training a neural network that translates simple synthetic images of a human character into realistic imagery. For training our networks, we first track the 3D motion of the person in the video using the template model and subsequently generate a synthetically rendered version of the video. These images are then used to train a conditional generative adversarial network that translates synthetic images of the 3D model into realistic imagery of the human. We evaluate our method for the reenactment of another person that is tracked to obtain the motion data, and show video results generated from artist-designed skeleton motion. Our results outperform the state of the art in learning-based human image synthesis.

This article presents a new computational framework for constructing 3D self-supporting surfaces with isotropic stress. Inspired by the self-supporting property of catenary and the fact that catenoid (the surface of revolution of the catenary curve) is a minimal surface, we discover the relation between 3D self-supporting surfaces and 4D minimal hypersurfaces (which are 3-manifolds). Lifting the problem into 4D allows us to convert gravitational forces into tensions and reformulate the equilibrium problem to total potential energy minimization, which can be solved using a variational method. We prove that the hyper-generatrix of a 4D minimal hyper-surface of revolution is a 3D self-supporting surface, implying that constructing a 3D self-supporting surface is equivalent to volume minimization. We show that the energy functional is simply the surface’s gravitational potential energy, which in turn can be converted into a surface reconstruction problem with mean curvature constraint. Armed with our theoretical findings, we develop an iterative algorithm to construct 3D self-supporting surfaces from triangle meshes. Our method guarantees convergence and can produce near-regular triangle meshes, thanks to a local mesh refinement strategy similar to centroidal Voronoi tessellation. It also allows users to tune the geometry via specifying either the zero potential surface or its desired volume. We also develop a finite element method to verify the equilibrium condition on 3D triangle meshes. The existing thrust network analysis methods discretize both geometry and material by approximating the continuous stress field through uniaxial singular stresses, making them an ideal tool for analysis and design of beam structures. In contrast, our method works on piecewise linear surfaces with continuous material. Moreover, our method does not require the 3D-to-2D projection, therefore it also works for both height and non-height fields.

We present a visualization approach for the analysis of CO2 bubble-induced attenuation in porous rock formations. As a basis for this, we introduce customized techniques to extract CO2 bubbles and their surrounding porous structure from X-ray computed tomography data (XCT) measurements. To understand how the structure of porous media influences the occurrence and the shape of formed bubbles, we automatically classify and relate them in terms of morphology and geometric features, and further directly support searching for promising porous structures. To allow for the meaningful direct visual comparison of bubbles and their structures, we propose a customized registration technique considering the bubble shape as well as its points of contact with the porous media surface. With our quantitative extraction of geometric bubble features, we further support the analysis as well as the creation of a physical model. We demonstrate that our approach was successfully used to answer several research questions in the domain, and discuss its high practical relevance to identify critical seismic characteristics of fluid-saturated rock that govern its capability to store CO2.

Digital cameras and mobile phones enable us to conveniently record precious moments. While digital image quality is constantly being improved, taking high-quality photos of digital screens still remains challenging because the photos are often contaminated with moiré patterns, a result of the interference between the pixel grids of the camera sensor and the device screen. Moiré patterns can severely damage the visual quality of photos. However, few studies have aimed to solve this problem. In this paper, we introduce a novel multiresolution fully convolutional network for automatically removing moiré patterns from photos. Since a moiré pattern spans over a wide range of frequencies, our proposed network performs a nonlinear multiresolution analysis of the input image before computing how to cancel moiré artefacts within every frequency band. We also create a large-scale benchmark dataset with 100,000+ image pairs for investigating and evaluating moiré pattern removal algorithms. Our network achieves state-of-the-art performance on this dataset in comparison to existing learning architectures for image restoration problems.

Sketching provides an intuitive user interface for communicating free form shapes. While human observers can easily envision the shapes they intend to communicate, replicating this process algorithmically requires resolving numerous ambiguities. Existing sketch-based modeling methods resolve these ambiguities by either relying on expensive user annotations or by restricting the modeled shapes to specific narrow categories. We present an approach for modeling generic freeform 3D surfaces from sparse, expressive 2D sketches that overcomes both limitations by incorporating convolution neural networks (CNN) into the sketch processing workflow. Given a 2D sketch of a 3D surface, we use CNNs to infer the depth and normal maps representing the surface. To combat ambiguity we introduce an intermediate CNN layer that models the dense curvature direction, or flow, field of the surface, and produce an additional output confidence map along with depth and normal. The flow field guides our subsequent surface reconstruction for improved regularity; the confidence map trained unsupervised measures ambiguity and provides a robust estimator for data fitting. To reduce ambiguities in input sketches users can refine their input by providing optional depth values at sparse points and curvature hints for strokes. Our CNN is trained on a large dataset generated by rendering sketches of various 3D shapes using non-photo-realistic line rendering (NPR) method that mimics human sketching of free-form shapes. We use the CNN model to process both single- and multi-view sketches. Using our multi-view framework users progressively complete the shape by sketching in different views, generating complete closed shapes. For each new view, the modeling is assisted by partial sketches and depth cues provided by surfaces generated in earlier views. The partial surfaces are fused into a complete shape using predicted confidence levels as weights. We validate our approach, compare it with previous methods and alternative structures, and evaluate its performance with various modeling tasks. The results demonstrate our method is a new approach for efficiently modeling freeform shapes with succinct but expressive 2D sketches.

We introduce C urve F usion , the first approach for high quality scanning of thin structures at interactive rates using a handheld RGBD camera. Thin filament-like structures are mathematically just 1D curves embedded in R 3 , and integration-based reconstruction works best when depth sequences (from the thin structure parts) are fused using the object's (unknown) curve skeleton. Thus, using the complementary but noisy color and depth channels, C urve F usion first automatically identifies point samples on potential thin structures and groups them into bundles , each being a group of a fixed number of aligned consecutive frames. Then, the algorithm extracts per-bundle skeleton curves using L 1 axes, and aligns and iteratively merges the L 1 segments from all the bundles to form the final complete curve skeleton. Thus, unlike previous methods, reconstruction happens via integration along a data-dependent fusion primitive , i.e., the extracted curve skeleton. We extensively evaluate C urve F usion on a range of challenging examples, different scanner and calibration settings, and present high fidelity thin structure reconstructions previously just not possible from raw RGBD sequences.

We present an automatic algorithm for subtractive manufacturing of freeform 3D objects using high-speed machining (HSM) via CNC. A CNC machine operates a cylindrical cutter to carve off material from a 3D shape stock, following a tool path, to "expose" the target object. Our method decomposes the input object's surface into a small number of patches each of which is fully accessible and machinable by the CNC machine, in continuous fashion, under a fixed cutter-object setup configuration. This is achieved by covering the input surface with a minimum number of accessible regions and then extracting a set of machinable patches from each accessible region. For each patch obtained, we compute a continuous, space-filling, and iso-scallop tool path which conforms to the patch boundary, enabling efficient carving with high-quality surface finishing. The tool path is generated in the form of connected Fermat spirals , which have been generalized from a 2D fill pattern for layered manufacturing to work for curved surfaces. Furthermore, we develop a novel method to control the spacing of Fermat spirals based on directional surface curvature and adapt the heat method to obtain iso-scallop carving. We demonstrate automatic generation of accessible and machinable surface decompositions and iso-scallop Fermat spiral carving paths for freeform 3D objects. Comparisons are made to tool paths generated by commercial software in terms of real machining time and surface quality.

This article presents a new method to compute a self-intersection free high-dimensional Euclidean embedding (SIFHDE 2 ) for surfaces and volumes equipped with an arbitrary Riemannian metric. It is already known that given a high-dimensional (high-d) embedding, one can easily compute an anisotropic Voronoi diagram by back-mapping it to 3D space. We show here how to solve the inverse problem, i.e., given an input metric, compute a smooth intersection-free high-d embedding of the input such that the pullback metric of the embedding matches the input metric. Our numerical solution mechanism matches the deformation gradient of the 3D → higher-d mapping with the given Riemannian metric. We demonstrate the applicability of our method, by using it to construct anisotropic Restricted Voronoi Diagram (RVD) and anisotropic meshing, that are otherwise extremely difficult to compute. In SIFHDE 2 -space constructed by our algorithm, difficult 3D anisotropic computations are replaced with simple Euclidean computations, resulting in an isotropic RVD and its dual mesh on this high-d embedding. Results are compared with the state-of-the-art in anisotropic surface and volume meshings using several examples and evaluation metrics.

A new method is presented for producing photo collages that preserve content correlation of photos. We use deep learning techniques to find correlation among given photos to facilitate their embedding on the canvas, and develop an efficient combinatorial optimization technique to make correlated photos stay close to each other. To make efficient use of canvas space, our method first extracts salient regions of photos and packs only these salient regions. We allow the salient regions to have arbitrary shapes, therefore yielding informative, yet more compact collages than by other similar collage methods based on salient regions. We present extensive experimental results, user study results, and comparisons against the state-of-the-art methods to show the superiority of our method.

Retrieving salient structure from textured images is an important but difficult problem in computer vision because texture, which can be irregular, anisotropic, non-uniform and complex, shares many of the same properties as structure. Observing that salient structure in a textured image should be piece-wise smooth, we present a method to retrieve such structures using an minimization of a modified form of the relative total variation metric. Thanks to the characteristics shared by texture and small structures, our method is effective at retrieving structure based on scale as well. Our method outperforms state-of-art methods in texture removal as well as scale-space filtering. We also demonstrate our method's ability in other applications such as edge detection, clip art compression artifact removal, and inverse half-toning.

The incorporation of subtitles in multimedia content plays an important role in communicating spoken content. For example, subtitles in the respective language are often preferred to expensive audio translation of foreign movies. The traditional representation of subtitles displays text centered at the bottom of the screen. This layout can lead to large distances between text and relevant image content, causing eye strain and even that we miss visual content. As a recent alternative, the technique of speaker-following subtitles places subtitle text in speech bubbles close to the current speaker. We conducted a controlled eye-tracking laboratory study (n = 40) to compare the regular approach (center-bottom subtitles) with content-sensitive, speaker-following subtitles. We compared different dialog-heavy video clips with the two layouts. Our results show that speaker-following subtitles lead to higher fixation counts on relevant image regions and reduce saccade length, which is an important factor for eye strain.

The availability of commodity 3D extruder pen allows direct drawing of 3D wire sculptures for novice users, enabling many novel applications such as intuitive spatial intelligence development for school students. However, the lack of spatial and structural cues among individual pen strokes makes the 3D drawing process challenging, which often leads to highly distorted and even incomplete wire sculptures. We present a mixed reality system, called `WireDraw', to immersively guide the 3D drawing for easy wire sculpturing. The system design is based on novel 3D drawing principles and the subsequent optimization, making the stroke sequence of the wire model drawable and easy to draw. On-the-fly edits on unsatisfactory strokes are also allowed for creative design. We demonstrate the effectiveness of our system by testing on a variety of wire models and a user study. The results show that the visual guidance provided by our system is extremely helpful for drawing high-quality wire sculptures.

Due to the development of 3D sensing and modeling techniques, the state-of-the-art mixed reality devices such as Microsoft Hololens have the ability of digitalizing the physical world. This unique feature bridges the gap between virtuality and reality and largely elevates the user experience. Unfortunately, the current solution only performs well if the virtual contents complement the real scene. It can easily cause visual artifacts when the reality needs to be modified due to the virtuality (e.g., remove real objects to offer more space for virtual objects), a common scenario in mixed reality applications such as room redecoration and environment design. We present a novel system, called emph{SceneCtrl}, that allows the user to interactively edit the real scene sensed by Hololens, such that the reality can be adapted to suit virtuality. Our proof-of-concept prototype employs scene reconstruction and understanding to enable efficient editing such as deleting, moving, and copying real objects in the scene. We also demonstrate emph{SceneCtrl} on a number of example scenarios in mixed reality, verifying the enhanced experience by resolving conflicts between virtuality and reality.

We present a novel superpixel generation algorithm based on a new definition of geodesic distance, called bilateral geodesic distance. In contrast to the traditional geodesic distance, the new bilateral geodesic distance of two pixels considers the distance between their positions as well as their color difference. Superpixel generation is essentially a problem of clustering image pixels with respect to a set of properly selected seeds. We first use an adaptive hexagonal subdivision method to determine the initial seed-based image gradient. Then, we use the bilateral geodesic distance to measure the similarity between the pixels and the seeds. We apply an improved fast marching method to generate superpixels’ contour regions with the expansion velocities dependent on a new gradient formulation that depends on the seeds’ properties. The experimental results indicate that our algorithm is not only much faster than the structure-based method, which uses conventional geodesic distance, but also outperforms the existing methods in terms of region compactness and region boundary regularity.

We present a video vectorization method that generates a video in vector representation from an input video in raster representation. A vector-based video representation offers the benefits of vector graphics, such as compactness and scalability. The vector video we generate is represented by a simplified tetrahedral control mesh over the spatial-temporal video volume, with color attributes defined at the mesh vertices. We present novel techniques for simplification and subdivision of a tetrahedral mesh to achieve high simplification ratio while preserving features and ensuring color fidelity. From an input raster video, our method is capable of generating a compact video in vector representation that allows a faithful reconstruction with low reconstruction errors.

Recent advances in 3D printing have made it easier to manufacture customized objects by ordinary users in an affordable manner, and therefore spurred high demand for more accessible methods for designing and fabricating 3D objects of various shapes and functionalities. In this paper we present a novel approach to model and fabricate surface-like objects composed of connected tiles, which can be used as objects in daily life, such as ornaments, covers, shades or handbags.

Sketch-based modeling provides a powerful paradigm for geometric modeling. Recent research had shown, sketch based modeling methods are most effective when targeting a specific family of surfaces. A large and growing arsenal of sketching tools is available for different types of geometries and different target user populations. Our work augments this arsenal with a new and powerful tool for modeling complex freeform shapes by sketching sparse 2D strokes; our method complements existing approaches in enabling the generation of surfaces with complex curvature patterns that are challenging to produce with existing methods. To model a desired surface patch with our technique, the user sketches the patch boundary as well as a small number of strokes representing the major bending directions of the shape. Our method uses this input to generate a curvature field that conforms to the user strokes and then uses this field to derive a freeform surface with the desired curvature pattern. To infer the surface from the strokes we first disambiguate the convex versus concave bending directions indicated by the strokes and estimate the surface bending magnitude along the strokes. We subsequently construct a curvature field based on these estimates, using a non-orthogonal 4-direction field coupled with a scalar magnitude field, and finally construct a surface whose curvature pattern reflects this field through an iterative sequence of simple linear optimizations. Our framework is well suited for single-view modeling, but also supports multi-view interaction, necessary to model complex shapes portions of which can be occluded in many views. It effectively combines multi-view inputs to obtain a coherent 3D shape. It runs at interactive speed allowing for immediate user feedback. We demonstrate the effectiveness of the proposed method through a large collection of complex examples created by both artists and amateurs. Our framework provides a useful complement to the existing sketch-based modeling methods.

Objects created by connecting and bending wires are common in furniture design, metal sculpting, wire jewelry, etc. Reconstructing such objects with traditional depth and image based methods is extremely difficult due to their unique characteristics such as lack of features, thin elements, and severe self-occlusions. We present a novel image-based method that reconstructs a set of continuous 3D wires used to create such an object, where each wire is composed of an ordered set of 3D curve segments. Our method exploits two main observations: simplicity - wire objects are often created using only a small number of wires, and smoothness - each wire is primarily smoothly bent with sharp features appearing only at joints or isolated points. In light of these observations, we tackle the challenging image correspondence problem across featureless wires by first generating multiple candidate 3D curve segments and then solving a global selection problem that balances between image and smoothness cues to identify the correct 3D curves. Next, we recover a decomposition of such curves into a set of distinct and continuous wires by formulating a multiple traveling salesman problem , which finds smooth paths, i.e. , wires, connecting the curves. We demonstrate our method on a wide set of real examples with varying complexity and present high-fidelity results using only 3 images for each object. We provide the source code and data for our work in the project website.

In this paper, we present a novel method on surface partition from the perspective of approximation theory. Different from previous shape proxies, the ellipsoidal variance proxy is proposed to penalize the partition results falling into disconnected parts. On its support, the Principle Component Analysis (PCA) based energy is developed for asymptotic cluster aspect ratio and size control. We provide the theoretical explanation on how the minimization of the PCA-based energy leads to the optimal asymptotic behavior for approximation. Moreover, we show the partitions on densely sampled triangular meshes converge to the theoretic expectations. To evaluate the effectiveness of surface approximation, polygonal/triangular surface remeshing results are generated. The experimental results demonstrate the high approximation quality of our method.

We present a decentralized algorithm for group-based coherent and reciprocal multi-agent navigation. In addition to generating collision-free trajectories for each agent, our approach is able to simulate macroscopic group movements and proxemic behaviors that result in coherent navigation. Our approach is general, makes no assumptions about the size or shape of the group, and can generate smooth trajectories for the agents. Furthermore, it can dynamically adapt to obstacles or the behavior of other agents. The additional overhead of generating proxemic group behaviors is relatively small and our approach can simulate hundreds of agents in real-time. We highlight its benefits on different benchmarks.

Abstract Bas‐relief is designed to provide 3D perception for the viewers under illumination. For the problem of bas‐relief generation from 3D object, most existing methods ignore the influence of illumination on bas‐relief appearance. In this paper, we propose a novel method that adaptively generate bas‐reliefs with respect to illumination conditions. Given a 3D object and its target appearance, our method finds an adaptive surface that preserves the appearance of the input. We validate our approach through a variety of applications. Experimental results indicate that the proposed approach is effective in producing bas‐reliefs with desired appearance under illumination.

Filigrees are thin patterns found in jewelry, ornaments and lace fabrics. They are often formed of repeated base elements manually composed into larger, delicate patterns. Digital fabrication simplifies the process of turning a virtual model of a filigree into a physical object. However, designing a virtual model of a filigree remains a time consuming and challenging task. The difficulty lies in tightly packing together the base elements while covering a target surface. In addition, the filigree has to be well connected and sufficiently robust to be fabricated. We propose a novel approach automating this task. Our technique covers a target surface with a set of input base elements, forming a filigree strong enough to be fabricated. We exploit two properties of filigrees to make this possible. First, as filigrees form delicate traceries they are well captured by their skeleton. This affords for a simpler definition of operators such as matching and deformation. Second, instead of seeking for a perfect packing of the base elements we relax the problem by allowing appearance preserving partial overlaps. We optimize a filigree by a stochastic search, further improved by a novel boosting algorithm that records and reuses good configurations discovered during the process. We illustrate our technique on a number of challenging examples reproducing filigrees on large objects, which we manufacture by 3D printing. Our technique affords for several user controls, such as the scale and orientation of the elements.

This article presents a new method to optimally partition a geometric domain with capacity constraints on the partitioned regions. It is an important problem in many fields, ranging from engineering to economics. It is known that a capacity-constrained partition can be obtained as a power diagram with the squared L2 metric. We present a method with super-linear convergence for computing optimal partition with capacity constraints that outperforms the state-of-the-art in an order of magnitude. We demonstrate the efficiency of our method in the context of three different applications in computer graphics and geometric processing: displacement interpolation of function distribution, blue-noise point sampling, and optimal convex decomposition of 2D domains. Furthermore, the proposed method is extended to capacity-constrained optimal partition with respect to general cost functions beyond the squared Euclidean distance.

Shape description and feature detection are fundamental problems in computer graphics and geometric modeling. Among many existing techniques, those based on geodesic distance have proven effective in providing intrinsic and discriminative shape descriptors. In this article we introduce a new intrinsic function for a three-dimensional (3D) shape and use it for shape description and geometric feature detection. Specifically, we introduce the intrinsic girth function (IGF) defined on a 2D closed surface. For a point p on the surface, the value of the IGF at p is the length of the shortest nonzero geodesic path starting and ending at p . The IGF is invariant under isometry, insensitive to mesh tessellations, and robust to surface noise. We propose a fast method for computing the IGF and discuss its applications to shape retrieval and detecting tips, tubes, and plates that are constituent parts of 3D objects.

Mosaics are widely used for surface decoration to produce appealing visual effects. We present a method for synthesizing digital surface mosaics with irregularly shaped tiles, which are a type of tiles often used for mosaics design. Our method employs both continuous optimization and combinatorial optimization to improve tile arrangement. In the continuous optimization step, we iteratively partition the base surface into approximate Voronoi regions of the tiles and optimize the positions and orientations of the tiles to achieve a tight fit. Combination optimization performs tile permutation and replacement to further increase surface coverage and diversify tile selection. The alternative applications of these two optimization steps lead to rich combination of tiles and high surface coverage. We demonstrate the effectiveness of our solution with extensive experiments and comparisons.

Three-dimensional symmetric tensor fields have a wide range of applications in solid and fluid mechanics. Recent advances in the (topological) analysis of 3D symmetric tensor fields focus on degenerate tensors which form curves. In this paper, we introduce a number of feature surfaces, such as neutral surfaces and traceless surfaces, into tensor field analysis, based on the notion of eigenvalue manifold. Neutral surfaces are the boundary between linear tensors and planar tensors, and the traceless surfaces are the boundary between tensors of positive traces and those of negative traces. Degenerate curves, neutral surfaces, and traceless surfaces together form a partition of the eigenvalue manifold, which provides a more complete tensor field analysis than degenerate curves alone. We also extract and visualize the isosurfaces of tensor modes, tensor isotropy, and tensor magnitude, which we have found useful for domain applications in fluid and solid mechanics. Extracting neutral and traceless surfaces using the Marching Tetrahedra method can cause the loss of geometric and topological details, which can lead to false physical interpretation. To robustly extract neutral surfaces and traceless surfaces, we develop a polynomial description of them which enables us to borrow techniques from algebraic surface extraction, a topic well-researched by the computer-aided design (CAD) community as well as the algebraic geometry community. In addition, we adapt the surface extraction technique, called A-patches, to improve the speed of finding degenerate curves. Finally, we apply our analysis to data from solid and fluid mechanics as well as scalar field analysis.

The medial axis transform has long been known as an intrinsic shape representation supporting a variety of shape analysis and synthesis tasks. However, for a given shape, it is hard to obtain its faithful, concise and stable medial axis, which hinders the application of the medial axis. In this paper, we introduce the medial mesh, a new discrete representation of the medial axis. A medial mesh is a 2D simplicial complex coupled with a radius function that provides a piecewise linear approximation to the medial axis. We further present an effective algorithm for computing a concise and stable medial mesh for a given shape. Our algorithm is quantitatively driven by a shape approximation error metric, and progressively simplifies an initial medial mesh by iteratively contracting edges until the approximation error reaches a predefined threshold. We further demonstrate the superior efficiency and accuracy of our method over existing methods for medial axis simplification.

Automatic speaker naming is the problem of localizing as well as identifying each speaking character in a TV/movie/live show video. This is a challenging problem mainly attributes to its multimodal nature, namely face cue alone is insufficient to achieve good performance. Previous multimodal approaches to this problem usually process the data of different modalities individually and merge them using handcrafted heuristics. Such approaches work well for simple scenes, but fail to achieve high performance for speakers with large appearance variations. In this paper, we propose a novel convolutional neural networks (CNN) based learning framework to automatically learn the fusion function of both face and audio cues. We show that without using face tracking, facial landmark localization or subtitle/transcript, our system with robust multimodal feature extraction is able to achieve state-of-the-art speaker naming performance evaluated on two diverse TV series. The dataset and implementation of our algorithm are publicly available online.

Although finding correspondences between similar images is an important problem in image processing, the existing algorithms cannot find accurate and dense correspondences in images with significant changes in lighting/transformation or with the non-rigid objects. This paper proposes a novel method for finding accurate and dense correspondences between images even in these difficult situations. Starting with the non-rigid dense correspondence algorithm [1] to generate an initial correspondence map, we propose a new geometric filter that uses cubic B-Spline surfaces to approximate the correspondence mapping functions for shared objects in both images, thereby eliminating outliers and noise. We then propose an iterative algorithm which enlarges the region containing valid correspondences. Compared with the existing methods, our method is more robust to significant changes in lighting, color, or viewpoint. Furthermore, we demonstrate how to extend our surface approximation method to video editing by first generating a reliable correspondence map between a given source frame and each frame of a video. The user can then edit the source frame, and the changes are automatically propagated through the entire video using the correspondence map. To evaluate our approach, we examine applications of unsupervised image recognition and video texture editing, and show that our algorithm produces better results than those from state-of-the-art approaches.

We introduce in this paper a new approach that conveniently converts conversational videos into comics with manga-style layout. With our approach, the manga-style layout of a comic page is achieved in a content-driven manner, and the main components, including panels and word balloons, that constitute a visually pleasing comic page are intelligently organized . Our approach extracts key frames on speakers by using a speaker detection technique such that word balloons can be placed near the corresponding speakers. We qualitatively measure the information contained in a comic page. With the initial layout automatically determined, the final comic page is obtained by maximizing such a measure and optimizing the parameters relating to the optimal display of comics. An efficient Markov chain Monte Carlo sampling algorithm is designed for the optimization. Our user study demonstrates that users much prefer our manga-style comics to purely Western style comics. Extensive experiments and comparisons against previous work also verify the effectiveness of our approach.

The medial axis transform (MAT) is an important shape representation for shape approximation, shape recognition, and shape retrieval. Despite years of research, there is still a lack of effective methods for efficient, robust and accurate computation of the MAT. We present an efficient method, called Q-MAT , that uses quadratic error minimization to compute a structurally simple, geometrically accurate, and compact representation of the MAT. We introduce a new error metric for approximation and a new quantitative characterization of unstable branches of the MAT, and integrate them in an extension of the well-known quadric error metric (QEM) framework for mesh decimation. Q-MAT is fast, removes insignificant unstable branches effectively, and produces a simple and accurate piecewise linear approximation of the MAT. The method is thoroughly validated and compared with existing methods for MAT computation.

We propose a new approach for automatic surfacing of 3D curve networks, a long standing computer graphics problem which has garnered new attention with the emergence of sketch based modeling systems capable of producing such networks. Our approach is motivated by recent studies suggesting that artist-designed curve networks consist of descriptive curves that convey intrinsic shape properties, and are dominated byrepresentative flow linesdesigned to convey the principal curvature lines on the surface. Studies indicate that viewers complete the intended surface shape by envisioning a surface whose curvature lines smoothly blend these flow-line curves. Following these observations we design a surfacing framework that automatically aligns the curvature lines of the constructed surface with the representative flow lines and smoothly interpolates these representative flow, or curvature directions while minimizing undesired curvature variation. Starting with an initial triangle mesh of the network, we dynamically adapt the mesh to maximize the agreement between the principal curvature direction field on the surface and a smoothflow fieldsuggested by the representative flow-line curves. Our main technical contribution is a framework for curvature-based surface modeling, that facilitates the creation of surfaces with prescribed curvature characteristics. We validate our method via visual inspection, via comparison to artist created and ground truth surfaces, as well as comparison to prior art, and confirm that our results are well aligned with the computed flow fields and with viewer perception of the input networks.

Mesh surfaces with planar hexagonal faces, what we refer to as PH meshes, offer an elegant way of paneling freeform architectural surfaces due to their node simplicity (i.e., valence-3 nodes) and naturally appealing layout. We investigate PH meshes to understand how the shape, size, and pattern of PH faces are constrained by surface geometry. This understanding enables us to develop an effective method for paneling freeform architectural surfaces with PH meshes. Our method first constructs an ideal triangulation of a given smooth surface, guided by surface geometry. We show that such an ideal triangulation leads to a Dupin-regular PH mesh via tangent duality on the surface. We have developed several novel and effective techniques for improving undesirable mesh layouts caused by singular behaviors of surface curvature. We compute support structures associated with PH meshes, including exact vertex offsets and approximate edge offsets, as demanded in panel manufacturing. The efficacy of our method is validated by a number of architectural examples.

Although the problem of k-area coverage has been intensively investigated for dense wireless sensor networks (WSNs), how to arrive at a k-coverage sensor deployment that optimizes certain objectives in relatively sparse WSNs still faces both theoretical and practical difficulties. Moreover, only a handful of centralized algorithms have been proposed to elevate 2-D area coverage to 3-D surface coverage. In this paper, we present a practical algorithm, i.e., the Autonomous dePlOyment for Load baLancing k-surface cOverage (APOLLO), to move sensor nodes toward k-surface coverage, aiming at minimizing the maximum sensing range required by the nodes. APOLLO enables purely autonomous node deployment as it only entails localized computations. We prove the termination of the algorithm and the (local) optimality of the output. We also show that our optimization objective is closely related to other frequently considered objectives for 2-D area coverage. Therefore, our practical algorithm design also contributes to the theoretical understanding of the 2-D k-area coverage problem. Finally, we use extensive simulation results to both confirm our theoretical claims and demonstrate the efficacy of APOLLO.

Abstract Shape analysis finds many important applications in shape understanding, matching and retrieval. Among the various shape analysis methods, spectral shape analysis aims to study the spectrum of the Laplace–Beltrami operator of some well‐designed shape‐dependent equations and obtain a spectral shape descriptor that can in turn be used for shape analysis purposes. The success of such approaches depends greatly on the discriminating power of a shape descriptor. On the other hand, the medial axis of a shape is widely known for its complete shape representation. It is sensitive to small perturbation of the boundary of a shape which often poses difficulty in its effective use for shape analysis. In this paper, we propose a new spectral shape descriptor, called the medial axis spectrum for 2D shapes, which directly applies spectral analysis to the medial axes of the shapes. We extend the Laplace–Beltrami operator onto the medial axis, and take the solution to an extended Laplacian eigenvalue problem defined on the axis as the medial axis spectrum. The medial axis spectrum is robust in the presence of shape boundary noise, and is invariant under rigid transformations, uniform scaling and isometry of the medial axis. We demonstrate these benefits of such a medial axis spectrum representation through extensive experiments. The medial axis spectrum is further used for 2D shape retrieval, and its superiority over previous work is shown by comparison.

Multiple videos may share a common foreground object, for instance a family member in home videos, or a leading role in various clips of a movie or TV series. In this paper, we present a novel method for co-segmenting the common foreground object from a group of video sequences. The issue was seldom touched on in the literature. Starting from over-segmentation of each video into Temporal Superpixels (TSPs), we first propose a new subspace clustering algorithm which segments the videos into consistent spatio-temporal regions with multiple classes, such that the common foreground has consistent labels across different videos. The subspace clustering algorithm exploits the fact that across different videos the common foreground shares similar appearance features, while motions can be used to better differentiate regions within each video, making accurate extraction of object boundaries easier. We further formulate video object co-segmentation as a Markov Random Field (MRF) model which imposes the constraint of foreground model automatically computed or specified with little user effort. The Quadratic Pseudo-Boolean Optimization (QPBO) is used to generate the results. Experiments show that this video co-segmentation framework can achieve good quality foreground extraction results without user interaction for those videos with unrelated background, and with only moderate user interaction for those videos with similar background. Comparisons with previous work also show the superiority of our approach.

Existing methods for surface quadrangulation cannot ensure accurate alignment with feature or boundary curves and tight control of local element size, which are important requirements in many numerical applications (e.g., FEA). Some methods rely on a prescribed direction field to guide quadrangulation for feature alignment, but such a direction field may conflict with a desired density field, thus making it difficult to control the element size. We propose a new spectral method that achieves both accurate feature curve alignment and tight control of local element size according to a given density field. Specifically, the following three technical contributions are made. First, to make the quadrangulation align accurately with feature curves or surface boundary curves, we introduce novel boundary conditions for wave-like functions that satisfy the Helmholtz equation approximately in the least squares sense. Such functions, called quasi-eigenfunctions , are computed efficiently as the solutions to a variational problem. Second, the mesh element size is effectively controlled by locally modulating the Laplace operator in the Helmholtz equation according to a given density field. Third, to improve robustness, we propose a novel scheme to minimize the vibration difference of the quasi-eigenfunction in two orthogonal directions. It is demonstrated by extensive experiments that our method outperforms previous methods in generating feature-aligned quadrilateral meshes with tight control of local elememt size. We further present some preliminary results to show that our method can be extended to generating hex-dominant volume meshes.

We propose a new method for improving the presentation of subtitles in video (e.g., TV and movies). With conventional subtitles, the viewer has to constantly look away from the main viewing area to read the subtitles at the bottom of the screen, which disrupts the viewing experience and causes unnecessary eyestrain. Our method places on-screen subtitles next to the respective speakers to allow the viewer to follow the visual content while simultaneously reading the subtitles. We use novel identification algorithms to detect the speakers based on audio and visual information. Then the placement of the subtitles is determined using global optimization. A comprehensive usability study indicated that our subtitle placement method outperformed both conventional fixed-position subtitling and another previous dynamic subtitling method in terms of enhancing the overall viewing experience and reducing eyestrain.

We present a novel artistic-verisimilitude driven system for watercolor rendering of images and photos. Our system achieves realistic simulation of a set of important characteristics of watercolor paintings that have not been well implemented before. Specifically, we designed several image filters to achieve: 1) watercolor-specified color transferring; 2) saliency-based level-of-detail drawing; 3) hand tremor effect due to human neural noise; and 4) an artistically controlled wet-in-wet effect in the border regions of different wet pigments. A user study indicates that our method can produce watercolor results of artistic verisimilitude better than previous filter-based or physical-based methods. Furthermore, our algorithm is efficient and can easily be parallelized, making it suitable for interactive image watercolorization.

In this paper, we present a novel approach for automatically creating the photo collage that assembles the interest regions of a given group of images naturally. Previous methods on photo collage are generally built upon a well-defined optimization framework, which computes all the geometric parameters and layer indices for input photos on the given canvas by optimizing a unified objective function. The complex nonlinear form of optimization function limits their scalability and efficiency. From the geometric point of view, we recast the generation of collage as a region partition problem such that each image is displayed in its corresponding region partitioned from the canvas. The core of this is an efficient power-diagram-based circle packing algorithm that arranges a series of circles assigned to input photos compactly in the given canvas. To favor important photos, the circles are associated with image importances determined by an image ranking process. A heuristic search process is developed to ensure that salient information of each photo is displayed in the polygonal area resulting from circle packing. With our new formulation, each factor influencing the state of a photo is optimized in an independent stage, and computation of the optimal states for neighboring photos are completely decoupled. This improves the scalability of collage results and ensures their diversity. We also devise a saliency-based image fusion scheme to generate seamless compositive collage. Our approach can generate the collages on nonrectangular canvases and supports interactive collage that allows the user to refine collage results according to his/her personal preferences. We conduct extensive experiments and show the superiority of our algorithm by comparing against previous methods.

Recognizing the category of a visual object remains a challenging computer vision problem. In this paper we develop a novel deep learning method that facilitates example-based visual object category recognition. Our deep learning architecture consists of multiple stacked layers and computes an intermediate representation that can be fed to a nearest-neighbor classifier. This intermediate representation is discriminative and structure-preserving. It is also capable of extracting essential characteristics shared by objects in the same category while filtering out nonessential differences among them. Each layer in our model is a nonlinear mapping, whose parameters are learned through two sequential steps that are designed to achieve the aforementioned properties. The first step computes a discrete mapping called supervised Laplacian Eigenmap. The second step computes a continuous mapping from the discrete version through nonlinear regression. We have extensively tested our method and it achieves state-of-the-art recognition rates on a number of benchmark datasets.

Blue noise sampling is an important component in many graphics applications, but existing techniques consider mainly the spatial positions of samples, making them less effective when handling problems with non-spatial features. Examples include biological distribution in which plant spacing is influenced by non-positional factors such as tree type and size, photon mapping in which photon flux and direction are not a direct function of the attached surface, and point cloud sampling in which the underlying surface is unknown a priori. These scenarios can benefit from blue noise sample distributions, but cannot be adequately handled by prior art. Inspired by bilateral filtering, we propose a bilateral blue noise sampling strategy. Our key idea is a general formulation to modulate the traditional sample distance measures, which are determined by sample position in spatial domain, with a similarity measure that considers arbitrary per sample attributes. This modulation leads to the notion of bilateral blue noise whose properties are influenced by not only the uniformity of the sample positions but also the similarity of the sample attributes. We describe how to incorporate our modulation into various sample analysis and synthesis methods, and demonstrate applications in object distribution, photon density estimation, and point cloud sub-sampling.

Masonry structures must be compressively self-supporting; designing such surfaces forms an important topic in architecture as well as a challenging problem in geometric modeling. Under certain conditions, a surjective mapping exists between a power diagram , defined by a set of 2D vertices and associated weights, and the reciprocal diagram that characterizes the force diagram of a discrete self-supporting network. This observation lets us define a new and convenient parameterization for the space of self-supporting networks. Based on it and the discrete geometry of this design space, we present novel geometry processing methods including surface smoothing and remeshing which significantly reduce the magnitude of force densities and homogenize their distribution.

This paper introduces a particle-based approach for anisotropic surface meshing. Given an input polygonal mesh endowed with a Riemannian metric and a specified number of vertices, the method generates a metric-adapted mesh. The main idea consists of mapping the anisotropic space into a higher dimensional isotropic one, called "embedding space". The vertices of the mesh are generated by uniformly sampling the surface in this higher dimensional embedding space, and the sampling is further regularized by optimizing an energy function with a quasi-Newton algorithm. All the computations can be re-expressed in terms of the dot product in the embedding space, and the Jacobian matrices of the mappings that connect different spaces. This transform makes it unnecessary to explicitly represent the coordinates in the embedding space, and also provides all necessary expressions of energy and forces for efficient computations. Through energy optimization, it naturally leads to the desired anisotropic particle distributions in the original space. The triangles are then generated by computing the Restricted Anisotropic Voronoi Diagram and its dual Delaunay triangulation. We compare our results qualitatively and quantitatively with the state-of-the-art in anisotropic surface meshing on several examples, using the standard measurement criteria.

Although the problem of k-area coverage has been intensively investigated for dense wireless sensor networks (WSNs), how to arrive at a k-coverage sensor deployment that optimizes certain objectives in relatively sparse WSNs still faces both theoretical and practical difficulties. In this paper, we present a practical algorithm LAACAD (Load balancing k-Area Coverage through Autonomous Deployment) to move sensor nodes toward k-area coverage, aiming at minimizing the maximum sensing range required by the nodes. LAACAD enables purely autonomous node deployment as it only entails localized computations. We prove the convergence of the algorithm, as well as the (local) optimality of the output. We also show that our optimization objective is closely related to other frequently considered objectives. Therefore, our practical algorithm design also contributes to the theoretical understanding of the k-area coverage problem. Finally, we use extensive simulation results both to confirm our theoretical claims and to demonstrate the efficacy of LAACAD.

Abstract We develop a novel isotropic remeshing method based onconstrained centroidal Delaunay mesh(CCDM), a generalization of centroidal patch triangulation from 2D to mesh surface. Our method starts with resampling an input mesh with a vertex distribution according to a user‐defined density function. The initial remeshing result is then progressively optimized by alternatively recovering the Delaunay mesh and moving each vertex to the centroid of its 1‐ring neighborhood. The key to making such simple iterations work is an efficient optimization framework that combines both local and global optimization methods. Our method is parameterization‐free, thus avoiding the metric distortion introduced by parameterization, and generating more well‐shaped triangles. Our method guarantees that the topology of surface is preserved without requiring geodesic information. We conduct various experiments to demonstrate the simplicity, efficacy, and robustness of the presented method.

Abstract Optimization of images with bad compositions has attracted increasing attention in recent years. Previous methods however seldomly consider image similarity when improving composition aesthetics. This may lead to significant content changes or bring large distortions, resulting in an unpleasant user experience. In this paper, we present a new algorithm for improving image composition aesthetics, while retaining faithful, as much as possible, to the original image content. Our method computes an improved image using a unified model of composition aesthetics and image similarity. The term of composition aesthetics obeys the rule of thirds and aims to enhance image composition. The similarity term in contrast penalizes image difference and distortion caused by composition adjustment. We use an edge‐based measure of structure similarity which nearly coincides with human visual perception to compare the optimized image with the original one. We describe an effective scheme to generate the optimized image with the objective model. Our algorithm is able to produce the recomposed images with minimal visual distortions in an elegant and user controllable manner. We show the superiority of our algorithm by comparing our results with those by previous methods.

Abstract Centroidal Voronoi Tessellation (CVT) of points has many applications in geometry processing, including re‐meshing and segmentation, to name but a few. In this paper, we generalize the CVT concept to graphs via a variational characterization. Given a graph and a 3D polygonal surface, our method optimizes the placement of the vertices of the graph in such a way that the graph segments best approximate the shape of the surface. We formulate the computation of CVT for graphs as a continuous variational problem, and present a simple, approximate method for solving this problem. Our method is robust in the sense that it is independent of degeneracies in the input mesh, such as skinny triangles, T‐junctions, small gaps or multiple connected components. We present some applications, to skeleton fitting and to shape segmentation.

Decomposing a volume into high-quality hexahedral cells is a challenging task in geometric modeling and computational geometry. Inspired by the use of cross field in quad meshing and the CubeCover approach in hex meshing, we present a complete all-hex meshing framework based on singularity-restricted field that is essential to induce a valid all-hex structure. Given a volume represented by a tetrahedral mesh, we first compute a boundary-aligned 3D frame field inside it, then convert the frame field to be singularity-restricted by our effective topological operations. In our all-hex meshing framework, we apply the CubeCover method to achieve the volume parametrization. For reducing degenerate elements appearing in the volume parametrization, we also propose novel tetrahedral split operations to preprocess singularity-restricted frame fields. Experimental results show that our algorithm generates high-quality all-hex meshes from a variety of 3D volumes robustly and efficiently.

We present a new method for modeling discrete constant mean curvature (CMC) surfaces, which arise frequently in nature and are highly demanded in architecture and other engineering applications. Our method is based on a novel use of the CVT ( centroidal Voronoi tessellation ) optimization framework. We devise a CVT-CMC energy function defined as a combination of an extended CVT energy and a volume functional. We show that minimizing the CVT-CMC energy is asymptotically equivalent to minimizing mesh surface area with a fixed volume, thus defining a discrete CMC surface. The CVT term in the energy function ensures high mesh quality throughout the evolution of a CMC surface in an interactive design process for form finding. Our method is capable of modeling CMC surfaces with fixed or free boundaries and is robust with respect to input mesh quality and topology changes. Experiments show that the new method generates discrete CMC surfaces of improved mesh quality over existing methods.

Blue noise point sampling is one of the core algorithms in computer graphics. In this paper, we present a new and versatile variational framework for generating point distributions with high-quality blue noise characteristics while precisely adapting to given density functions. Different from previous approaches based on discrete settings of capacity-constrained Voronoi tessellation, we cast the blue noise sampling generation as a variational problem with continuous settings. Based on an accurate evaluation of the gradient of an energy function, an efficient optimization is developed which delivers significantly faster performance than the previous optimization-based methods. Our framework can easily be extended to generating blue noise point samples on manifold surfaces and for multi-class sampling. The optimization formulation also allows us to naturally deal with dynamic domains, such as deformable surfaces, and to yield blue noise samplings with temporal coherence. We present experimental results to validate the efficacy of our variational framework. Finally, we show a variety of applications of the proposed methods, including nonphotorealistic image stippling, color stippling, and blue noise sampling on deformable surfaces.

Centroidal Voronoi Tessellation (CVT) is a widely used geometric structure in applications including mesh generation, vector quantization and image processing. Global optimization of the CVT function is important in these applications. With numerical evidences, we show that the CVT function is highly nonconvex and has many local minima and therefore the global optimization of the CVT function is nontrivial. We apply the method of Monte Carlo with Minimization (MCM) to optimizing the CVT function globally and demonstrate its efficacy in producing much improved results compared with two other global optimization methods.

Abstract Depth‐of‐field is one of the most crucial rendering effects for synthesizing photorealistic images. Unfortunately, this effect is also extremely costly. It can take hundreds to thousands of samples to achieve noise‐free results using Monte Carlo integration. This paper introduces an efficient adaptive depth‐of‐field rendering algorithm that achieves noise‐free results using significantly fewer samples. Our algorithm consists of two main phases: adaptive sampling and image reconstruction. In the adaptive sampling phase, the adaptive sample density is determined by a ‘blur‐size’ map and ‘pixel‐variance’ map computed in the initialization. In the image reconstruction phase, based on the blur‐size map, we use a novel multiscale reconstruction filter to dramatically reduce the noise in the defocused areas where the sampled radiance has high variance. Because of the efficiency of this new filter, only a few samples are required. With the combination of the adaptive sampler and the multiscale filter, our algorithm renders near‐reference quality depth‐of‐field images with significantly fewer samples than previous techniques.

The most important guiding principle in computational methods for freeform architecture is the balance between cost efficiency on the one hand, and adherence to the design intent on the other. Key issues are the simplicity of supporting and connecting elements as well as repetition of costly parts. This paper proposes so-called circular arc structures as a means to faithfully realize freeform designs without giving up smooth appearance. In contrast to non-smooth meshes with straight edges where geometric complexity is concentrated in the nodes, we stay with smooth surfaces and rather distribute complexity in a uniform way by allowing edges in the shape of circular arcs. We are able to achieve the simplest possible shape of nodes without interfering with known panel optimization algorithms. We study remarkable special cases of circular arc structures which possess simple supporting elements or repetitive edges, we present the first global approximation method for principal patches, and we show an extension to volumetric structures for truly three-dimensional designs.

Centroidal Voronoi tessellations (CVT) are widely used in computational science and engineering. The most commonly used method is Lloyd's method, and recently the L-BFGS method is shown to be faster than Lloyd's method for computing the CVT. However, these methods run on the CPU and are still too slow for many practical applications. We present techniques to implement these methods on the GPU for computing the CVT on 2D planes and on surfaces, and demonstrate significant speedup of these GPU-based methods over their CPU counterparts. For CVT computation on a surface, we use a geometry image stored in the GPU to represent the surface for computing the Voronoi diagram on it. In our implementation a new technique is proposed for parallel regional reduction on the GPU for evaluating integrals over Voronoi cells.

Abstract Updating a Delaunay triangulation when data points are slightly moved is the bottleneck of computation time in variational methods for mesh generation and remeshing. Utilizing the connectivity coherence between two consecutive Delaunay triangulations for computation speedup is the key to solving this problem. Our contribution is an effective filtering technique that confirms most bi‐cells whose Delaunay connectivities remain unchanged after the points are perturbed. Based on bi‐cell flipping, we present an efficient algorithm for updating two‐dimensional and three‐dimensional Delaunay triangulations of dynamic point sets. Experimental results show that our algorithm outperforms previous methods.

This paper presents a reconstruction pipeline for recovering branching structure of trees from laser scanned data points. The process is made up of two main blocks: segmentation and reconstruction. Based on a variational k-means clustering algorithm, cylindrical components and ramified regions of data points are identified and located. An adjacency graph is then built from neighborhood information of components. Simple heuristics allow us to extract a skeleton structure and identify branches from the graph. Finally, a B-spline model is computed to give a compact and accurate reconstruction of the branching system.

Abstract We propose a new isotropic remeshing method, based onCentroidal Voronoi Tessellation (CVT). Constructing CVT requires to repeatedly computeRestricted Voronoi Diagram (RVD), defined as the intersection between a 3D Voronoi diagram and an input mesh surface. Existing methods use some approximations of RVD. In this paper, we introduce an efficient algorithm that computes RVD exactly and robustly. As a consequence, we achieve better remeshing quality than approximation‐based approaches, without sacrificing efficiency. Our method for RVD computation uses a simple procedure and akd‐tree to quickly identify and compute the intersection of each triangle face with its incident Voronoi cells. Its time complexity isO(mlogn), wherenis the number of seed points andmis the number of triangles of the input mesh. Fast convergence of CVT is achieved using a quasi‐Newton method, which proved much faster than Lloyd's iteration. Examples are presented to demonstrate the better quality of remeshing results with our method than with the state‐of‐art approaches.

Centroidal Voronoi tessellation (CVT) is a particular type of Voronoi tessellation that has many applications in computational sciences and engineering, including computer graphics. The prevailing method for computing CVT is Lloyd's method, which has linear convergence and is inefficient in practice. We develop new efficient methods for CVT computation and demonstrate the fast convergence of these methods. Specifically, we show that the CVT energy function has 2nd order smoothness for convex domains with smooth density, as well as in most situations encountered in optimization. Due to the 2nd order smoothness, it is possible to minimize the CVT energy functions using Newton-like optimization methods and expect fast convergence. We propose a quasi-Newton method to compute CVT and demonstrate its faster convergence than Lloyd's method with various numerical examples. It is also significantly faster and more robust than the Lloyd-Newton method, a previous attempt to accelerate CVT. We also demonstrate surface remeshing as a possible application.

Abstract Various methods have been proposed for fitting subdivision surfaces to different forms of shape data (e.g., dense meshes or point clouds), but none of these methods effectively deals with shapes with sharp features, that is, creases, darts and corners. We present an effective method for fitting a Loop subdivision surface to a dense triangle mesh with sharp features. Our contribution is a new exact evaluation scheme for the Loop subdivision with all types of sharp features, which enables us to compute a fitting Loop subdivision surface for shapes with sharp features in an optimization framework. With an initial control mesh obtained from simplifying the input dense mesh using QEM, our fitting algorithm employs an iterative method to solve a nonlinear least squares problem based on the squared distances from the input mesh vertices to the fitting subdivision surface. This optimization framework depends critically on the ability to express these distances as quadratic functions of control mesh vertices using our exact evaluation scheme near sharp features. Experimental results are presented to demonstrate the effectiveness of the method.

Motivated by applications in architecture and manufacturing, we discuss the problem of covering a freeform surface by single curved panels. This leads to the new concept of semi-discrete surface representation, which constitutes a link between smooth and discrete surfaces. The basic entity we are working with is the developable strip model. It is the semi-discrete equivalent of a quad mesh with planar faces, or a conjugate parametrization of a smooth surface. We present a B-spline based optimization framework for efficient computing with D-strip models. In particular we study conical and circular models, which semi-discretize the network of principal curvature lines, and which enjoy elegant geometric properties. Together with geodesic models and cylindrical models they offer a rich source of solutions for surface panelization problems.

Due to its minimal twist, the rotation minimizing frame (RMF) is widely used in computer graphics, including sweep or blending surface modeling, motion design and control in computer animation and robotics, streamline visualization, and tool path planning in CAD/CAM. We present a novel simple and efficient method for accurate and stable computation of RMF of a curve in 3D. This method, called the double reflection method , uses two reflections to compute each frame from its preceding one to yield a sequence of frames to approximate an exact RMF. The double reflection method has the fourth order global approximation error, thus it is much more accurate than the two currently prevailing methods with the second order approximation error—the projection method by Klok and the rotation method by Bloomenthal, while all these methods have nearly the same per-frame computational cost. Furthermore, the double reflection method is much simpler and faster than using the standard fourth order Runge-Kutta method to integrate the defining ODE of the RMF, though they have the same accuracy. We also investigate further properties and extensions of the double reflection method, and discuss the variational principles in design moving frames with boundary conditions, based on RMF.

Coarse piecewise linear approximation of surfaces causes undesirable polygonal appearance of silhouettes. We present an efficient method for smoothing the silhouettes of coarse triangle meshes using efficient 3D curve reconstruction and simple local re-meshing. It does not assume the availability of a fine mesh and generates only moderate amount of additional data at run time. Furthermore, polygonal feature edges are also smoothed in a unified framework. Our method is based on a novel interpolation scheme over silhouette triangles and this ensures that smooth silhouettes are faithfully reconstructed and always change continuously with respect to continuous movement of the view point or objects. We speed up computation with GPU assistance to achieve real-time rendering of coarse meshes with the smoothed silhouettes. Experiments show that this method outperforms previous methods for silhouette smoothing.

Abstract We propose a variational approach to computing an optimal segmentation of a 3D shape for computing a union of tight bounding volumes. Based on an affine invariant measure of e‐tightness, the resemblance to ellipsoid, a novel functional is formulated that governs an optimization process to obtain a partition with multiple components. Refinement of segmentation is driven by application‐specific error measures, so that the final bounding volume meets pre‐specified user requirement. We present examples to demonstrate the effectiveness of our method and show that it works well for computing ellipsoidal bounding volumes as well as oriented bounding boxes.

The geometric challenges in the architectural design of freeform shapes come mainly from the physical realization of beams and nodes. We approach them via the concept of parallel meshes, and present methods of computation and optimization. We discuss planar faces, beams of controlled height, node geometry, and multilayer constructions. Beams of constant height are achieved with the new type of edge offset meshes. Mesh parallelism is also the main ingredient in a novel discrete theory of curvatures. These methods are applied to the construction of quadrilateral, pentagonal and hexagonal meshes, discrete minimal surfaces, discrete constant mean curvature surfaces, and their geometric transforms. We show how to design geometrically optimal shapes, and how to find a meaningful meshing and beam layout for existing shapes.

We present a complete framework for computing a subdivision surface to approximate unorganized point sample data, which is a separable nonlinear least squares problem. We study the convergence and stability of three geometrically-motivated optimization schemes and reveal their intrinsic relations with standard methods for constrained nonlinear optimization. A commonly-used method in graphics, called point distance minimization, is shown to use a variant of the gradient descent step and thus has only linear convergence. The second method, called tangent distance minimization, which is well-known in computer vision, is shown to use the Gauss-Newton step, and thus demonstrates near quadratic convergence for zero residual problems but may not converge otherwise. Finally, we show that an optimization scheme called squared distance minimization, recently proposed by Pottmann et al., can be derived from the Newton method. Hence, with proper regularization, tangent distance minimization and squared distance minimization are more efficient than point distance minimization. We also investigate the effects of two step size control methods -- Levenberg-Marquardt regularization and the Armijo rule -- on the convergence stability and efficiency of the above optimization schemes.

Computing a curve to approximate data points is a problem encountered frequently in many applications in computer graphics, computer vision, CAD/CAM, and image processing. We present a novel and efficient method, called squared distance minimization (SDM), for computing a planar B-spline curve, closed or open, to approximate a target shape defined by a point cloud , that is, a set of unorganized, possibly noisy data points. We show that SDM significantly outperforms other optimization methods used currently in common practice of curve fitting. In SDM, a B-spline curve starts from some properly specified initial shape and converges towards the target shape through iterative quadratic minimization of the fitting error. Our contribution is the introduction of a new fitting error term, called the squared distance (SD) error term , defined by a curvature-based quadratic approximant of squared distances from data points to a fitting curve. The SD error term faithfully measures the geometric distance between a fitting curve and a target shape, thus leading to faster and more stable convergence than the point distance (PD) error term, which is commonly used in computer graphics and CAGD, and the tangent distance (TD) error term, which is often adopted in the computer vision community. To provide a theoretical explanation of the superior performance of SDM, we formulate the B-spline curve fitting problem as a nonlinear least squares problem and conclude that SDM is a quasi-Newton method which employs a curvature-based positive definite approximant to the true Hessian of the objective function. Furthermore, we show that the method based on the TD error term is a Gauss-Newton iteration, which is unstable for target shapes with high curvature variations, whereas optimization based on the PD error term is the alternating method that is known to have linear convergence.

Collision detection and avoidance are important in robotics. Compared with commonly used circular disks, elliptic disks provide a more compact shape representation for robots or other vehicles confined to move in the plane. Furthermore, elliptic disks allow a simpler analytic representation than rectangular boxes, which makes it easier to perform continuous collision detection (CCD). We shall present a fast and accurate method for CCD between two moving elliptic disks, which avoids any need to sample the time domain of the motion, thus avoiding the possibility of missing collisions between time samples. Based on some new algebraic conditions on the separation of two ellipses, we reduce collision detection for two moving ellipses to the problem of detecting real roots of a univariate equation, which is the discriminant of the characteristic polynomial of the two ellipses. Several techniques are investigated for robust and accurate processing of this univariate equation for two classes of commonly used motions: planar cycloidal motions and planar rational motions. Experimental results demonstrate the efficiency, accuracy, and robustness of our method.

We present a novel algorithm based on least-squares minimization to approximate point cloud data in 2D plane with a smooth B-spline curve. The point cloud data may represent an open curve with self intersection and sharp corner. Unlike other existing methods, such as the moving least-squares method and the principle curve method, our algorithm does not need a thinning process. The idea of our algorithm is intuitive and simple - we make a B-spline curve grow along the tangential directions at its two end-points following local geometry of point clouds. Our algorithm generates appropriate control points of the fitting B-spline curve in the least squares sense. Although presented for the 2D case, our method can be extended in a straightforward manner to fitting data points by a B-spline curve in higher dimensions

We study the reconstruction of smooth surfaces from point clouds. We use a new squared distance error term in optimization to fit a subdivision surface to a set of unorganized points, which defines a closed target surface of arbitrary topology. The resulting method is based on the framework of squared distance minimization (SDM) proposed by Pottmann et al. Specifically, with an initial subdivision surface having a coarse control mesh as input, we adjust the control points by optimizing an objective function through iterative minimization of a quadratic approximant of the squared distance function of the target shape. Our experiments show that the new method (SDM) converges much faster than the commonly used optimization method using the point distance error function, which is known to have only linear convergence. This observation is further supported by our recent result that SDM can be derived from the Newton method with necessary modifications to make the Hessian positive definite and the fact that the Newton method has quadratic convergence.

We present a new method for converting a photo or image to a synthesized painting following the painting style of an example painting. Treating painting styles of brush strokes as sample textures, we reduce the problem of learning an example painting to a texture synthesis problem. The proposed method uses a hierarchical patch-based approach to the synthesis of directional textures. The key features of our method are: 1) Painting styles are represented as one or more blocks of sample textures selected by the user from the example painting; 2) image segmentation and brush stroke directions defined by the medial axis are used to better represent and communicate shapes and objects present in the synthesized painting; 3) image masks and a hierarchy of texture patches are used to efficiently synthesize high-quality directional textures. The synthesis process is further accelerated through texture direction quantization and the use of Gaussian pyramids. Our method has the following advantages: First, the synthesized stroke textures can follow a direction field determined by the shapes of regions to be painted. Second, the method is very efficient; the generation time of a synthesized painting ranges from a few seconds to about one minute, rather than hours, as required by other existing methods, on a commodity PC. Furthermore, the technique presented here provides a new and efficient solution to the problem of synthesizing a 2D directional texture. We use a number of test examples to demonstrate the efficiency of the proposed method and the high quality of results produced by the method.

We present new results on classifying the morphology of the nonsingular intersection curve of two quadrics by studying the roots of the characteristic equation, or the discriminant, of the pencil spanned by the two quadrics. The morphology of a nonsingular algebraic curve means the structural (or topological) information about the curve, such as the number of disjoint connected components of the curve in P/spl Ropf//sup 3/ (the 3D real projective space), and whether a particular component is a compact set in any affine realization of P/spl Ropf//sup 3/. For example, we show that two quadrics intersect along a nonsingular space quartic curve in P/spl Ropf//sup 3/ with one connected component if and only if their characteristic equation has two distinct real roots and a pair of complex conjugate roots. Since the number of the real roots of the characteristic equation can be counted robustly with exact arithmetic, our results can be used to obtain structural information reliably before computing the parameterization of the intersection curve; thus errors in the subsequent computation that is most likely done using floating point arithmetic will not lead to erroneous topological classification of the intersection curve. The key technique used to prove our results is to reduce two quadrics into simple forms using a projective transformation, a technique equivalent to the simultaneous block diagonalization of two real symmetric matrices, a topic that has been studied in matrix algebra.

We present an algorithm for computing the distance between two free-form surfaces. Using line geometry, the distance computation is reformulated as a simple instance of a surface-surface intersection problem, which leads to low-dimensional root finding in a system of equations. This approach produces an efficient algorithm for computing the distance between two ellipsoids, where the problem is reduced to finding a specific solution in a system of two equations in two variables. Similar algorithms can be designed for computing the distance between an ellipsoid and a simple surface (such as cylinder cone, or torus). In an experimental implementation (on a 500 MHz Windows PC), the distance between two ellipsoids was computed in less than 0.3 msec on average; and the distance between an ellipsoid and a simple convex surface was computed in less than 0.15 msec on average.

The problem of collision detection is fundamental to inter-active applications such as computer animation and virtual environments. In these fields, prompt recognition of possible impacts is important for computing real-time response. We present a simple exact collision detection algorithm for convex polytopes. The algorithm finds quickly a separating plane between two polytopes if they are non-colliding, or else reports collision if it cannot possibly find a separating plane. In the case of non-collision, the separating plane found for one time frame is cached as a witness for the next time frame, an idea borrowed from [10]; this use of time coherence further speeds up the algorithm in dynamic applications. Both temporal and geometric coherences are exploited to make this algorithm run in expected constant time empirically.