With the growth of 3D applications and the rapid increase in sensor-collected 3D point cloud data, there is a rising demand for efficient compression algorithms. Most existing learning-based compression methods handle geometry and color attributes separately, treating them as distinct tasks, making these methods challenging to apply directly to point clouds with colors. Besides, the limited capacities of training datasets also limit their generalizability across points with different distributions. In this work, we introduce a test-time unified geometry and color compression framework for 3D point clouds. Instead of training a compression model based on specific datasets, we adapt a pre-trained generative diffusion model to compress original colored point clouds into sparse sets, termed 'seeds', using prompt tuning. Decompression is then achieved through multiple denoising steps with separate sampling processes. Experiments on objects and indoor scenes demonstrate that our method has superior performances compared to existing baselines for the compression of geometry and color. Our codes would be released at https://github.com/Tianxinhuang/DiffCom.git.

Recent implicit and primitive-based radiance field methods, such as NeRF and 3DGS, have demonstrated impressive capabilities in novel view synthesis from multi-view images. However, their custom representations are often incompatible with conventional graphics pipelines, limiting their application in areas like editing, relighting, physics simulation, and particle effects. Additionally, their volume rendering approach requires alpha-blending multiple colors per pixel, which slows down rendering. Traditional differentiable rendering methods, while offering higher compatibility and speed, rely on known mesh topology, making them unsuitable for complex scene-level reconstruction. To address these limitations, we introduce a novel end-toend optimization process of disjoint opaque triangles, natively compatible with standard graphics engines for a wide range of applications. To enable gradient-based optimization over the highly non-differentiable rasterization process, we employ a$2 D$SDF approximation and a two-layer occlusion approximation. We also incorporate density controls to ensure detailed and complete scene reconstructions. Our paper tackles the challenging end-to-end optimization of scene-level novel view synthesis with opaque representation only. Our approach achieves over 1000 FPS rendering on a single desktop GPU, providing high compatibility and similar novel view synthesis quality to existing methods.

Condensed datasets offer a compact representation of larger datasets, but training models directly on them or using them to enhance model performance through knowledge distillation (KD) can result in suboptimal outcomes due to limited information. To address this, we propose a method that expands condensed datasets using model inversion, a technique for generating synthetic data based on the impressions of a pre-trained model on its training data. This approach is particularly well-suited for KD scenarios, as the teacher model is already pre-trained and retains knowledge of the original training data. By creating synthetic data that complements the condensed samples, we enrich the training set and better approximate the underlying data distribution, leading to improvements in student model accuracy during knowledge distillation. Our method demonstrates significant gains in KD accuracy compared to using condensed datasets alone and outperforms standard model inversion-based KD methods by up to 11.4% across various datasets and model architectures. Importantly, it remains effective even when using as few as one condensed sample per class, and can also enhance performance in few-shot scenarios where only limited real data samples are available.

3D hand pose estimation that involves accurate estimation of 3D human hand keypoint locations is crucial for many human-computer interaction applications such as augmented reality. However, this task poses significant challenges due to self-occlusion of the hands and occlusions caused by interactions with objects. In this paper, we propose HandMCM to address these challenges. Our HandMCM is a novel method based on the powerful state space model (Mamba). By incorporating modules for local information injection/filtering and correspondence modeling, the proposed correspondence Mamba effectively learns the highly dynamic kinematic topology of keypoints across various occlusion scenarios. Moreover, by integrating multi-modal image features, we enhance the robustness and representational capacity of the input, leading to more accurate hand pose estimation. Empirical evaluations on three benchmark datasets demonstrate that our model significantly outperforms current state-of-the-art methods, particularly in challenging scenarios involving severe occlusions. These results highlight the potential of our approach to advance the accuracy and reliability of 3D hand pose estimation in practical applications.

While natural language is commonly used to guide embodied agents, the inherent ambiguity and verbosity of language often hinder the effectiveness of language-guided navigation in complex environments. To this end, we propose Visual Prompt Navigation (VPN), a novel paradigm that guides agents to navigate using only user-provided visual prompts within 2D top-view maps. This visual prompt primarily focuses on marking the visual navigation trajectory on a top-down view of a scene, offering intuitive and spatially grounded guidance without relying on language instructions. It is more friendly for non-expert users and reduces interpretive ambiguity. We build VPN tasks in both discrete and continuous navigation settings, constructing two new datasets, R2R-VP and R2R-CE-VP, by extending existing R2R and R2R-CE episodes with corresponding visual prompts. Furthermore, we introduce VPNet, a dedicated baseline network to handle the VPN tasks, with two data augmentation strategies: view-level augmentation (altering initial headings and prompt orientations) and trajectory-level augmentation (incorporating diverse trajectories from large-scale 3D scenes), to enhance navigation performance. Extensive experiments evaluate how visual prompt forms, top-view map formats, and data augmentation strategies affect the performance of visual prompt navigation.

Minimal parametrization of 3D lines plays a critical role in camera localization and structural mapping. Existing representations in robotics and computer vision predominantly handle independent lines, overlooking structural regularities such as sets of parallel lines that are pervasive in man-made environments. This paper introduces RiemanLine, a unified minimal representation for 3D lines formulated on Riemannian manifolds that jointly accommodates both individual lines and parallel-line groups. Our key idea is to decouple each line landmark into global and local components: a shared vanishing direction optimized on the unit sphere, and scaled normal vectors constrained on orthogonal subspaces, enabling compact encoding of structural regularities. For n parallel lines, the proposed representation reduces the parameter space from 4n (orthonormal form) to 2n+2, naturally embedding parallelism without explicit constraints. We further integrate this parameterization into a factor graph framework, allowing global direction alignment and local reprojection optimization within a unified manifold-based bundle adjustment. Extensive experiments on ICL-NUIM, TartanAir, and synthetic benchmarks demonstrate that our method achieves significantly more accurate pose estimation and line reconstruction, while reducing parameter dimensionality and improving convergence stability.

DETR-based 3D detectors have recently emerged as a popular alternative to voting- and voxel-based methods, which offer end-to-end set prediction without handcrafted priors or voxelization. However, they remain unexplored under semi-supervision, where the scarcity of annotated 3D data impedes their widespread adoption. In this work, we present Semi-3DETR, the first framework to systematically adapt DETR to semi-supervised 3D object detection by addressing challenges unique to 3D. Compared to 2D Semi-DETR, semi-supervised 3D DETR faces amplified issues of fragile volumetric pseudo-labels, unstable query alignment, and noisy bipartite matching. Our Semi-3DETR mitigates these issues by introducing three core components: Robust Pseudo-Label Denoising (RPLD) to filter and refine volumetric pseudo-labels against orientation and depth errors, Query Alignment Consistency (QAC) to stabilize teacher–student query correspondence under 3D transformations, and a Hybrid Matching Strategy (HMS) to balance one-to-one and one-to-many assignments under noisy supervision. We further adopt a softmax classifier to enforce class exclusivity and improve pseudo-label reliability in semantically ambiguous 3D categories. Extensive experiments on ScanNet and SUN RGB-D demonstrate the feasibility of our Semi-3DETR with promising results compared to fully supervised and semi-supervised baselines. The source code will be released upon paper acceptance.

Recently, the integration of the efficient feed-forward scheme into 3D Gaussian Splatting (3DGS) has been actively explored. However, most existing methods focus on sparse view reconstruction of small regions and cannot produce eligible whole-scene reconstruction results in terms of either quality or efficiency. In this paper, we propose FreeSplat++, which focuses on extending the generalizable 3DGS to become an alternative approach to large-scale indoor whole-scene reconstruction, which has the potential of significantly accelerating the reconstruction speed and improving the geometric accuracy. To facilitate whole-scene reconstruction, we initially propose the Low-cost Cross-View Aggregation framework to efficiently process extremely long input sequences. Subsequently, we introduce a carefully designed pixel-wise triplet fusion method to incrementally aggregate the overlapping 3D Gaussian primitives from multiple views, adaptively reducing their redundancy. Furthermore, given the fused 3DGS primitives with accumulated weights after the fusion step, we propose a weighted floater removal strategy that can effectively reduce floaters, which serves as an explicit depth fusion approach that is tailored for generalizable 3DGS methods and becomes crucial in whole-scene reconstruction. After the feed-forward reconstruction of 3DGS primitives, we investigate a depth-regularized per-scene fine-tuning process. Leveraging the dense, multi-view consistent depth maps obtained during the feed-forward prediction phase for an extra constraint, we refine the entire scene's 3DGS primitive to enhance rendering quality while preserving geometric accuracy. Extensive experiments confirm that our FreeSplat++ significantly outperforms existing generalizable 3DGS methods, especially in whole scene reconstructions. Compared to conventional per-scene optimized 3DGS approaches, our method with depth-regularized per-scene fine-tuning demonstrates substantial improvements in reconstruction accuracy and a notable reduction in training time.

This paper presents Particle Rendering (PR), a new implicit rendering approach that extends Neural Radiance Fields (NeRF) by incorporating incident light along with traditional outgoing light modeling. In our framework, a 3D scene consists of a mass of particles, each offering a deeper understanding of light interactions by reflecting and emitting light in all directions. Our methodology involves a three-phase training pipeline: 1) Estimating the outgoing light field through a NeRF model; 2) Distilling the incident light field. A simple metric is introduced to assess the quality of the ray for better supervision; 3) Implicit rendering. We propose an implicit method to aggregate incident and outgoing fields that leverages Multilayer Perceptrons (MLP) to directly infer final pixel values, thus avoiding the limitation of traditional physically-based rendering techniques. The effectiveness of PR is demonstrated through state-of-the-art results in various challenging indoor and outdoor scenes, emphasizing its capability to handle complex lighting and reflective materials.

Dynamic neural radiance fields (dynamic NeRFs) have achieved remarkable successes in synthesizing novel views for 3D dynamic scenes. Traditional approaches typically necessitate full video sequences for the training phase prior to the synthesis of new views, akin to replaying a recording of a dynamic 3D event. In contrast, on-the-fly training allows for the immediate processing and rendering of dynamic scenes without the need for pre-training on full sequences, offering a more flexible and time-efficient solution for dynamic scene rendering tasks. In this paper, we propose a highly efficient on-the-fly training algorithm for dynamic NeRFs, named OD-NeRF. To accelerate the training process, our method minimizes the training required for the model at each frame by using: 1) a NeRF model conditioned on multi-view projected colors, which exhibits superior generalization across multiple frames with minimal training, and 2) a transition and update algorithm that leverages the occupancy grid from the last frame to sample efficiently at the current frame. Our algorithm can achieve an interactive training speed of 10FPS on synthetic dynamic scenes on-the-fly, and a 3×-9× training speed-up compared to the state-of-the-art on-the-fly NeRF on real-world dynamic scenes.

Recent advances in auto-regressive transformers have revolutionized generative modeling across different domains, from language processing to visual generation, demonstrating remarkable capabilities. However, applying these advances to 3D generation presents three key challenges: the unordered nature of 3D data conflicts with sequential next-token prediction paradigm, conventional vector quantization approaches incur substantial compression loss when applied to 3D meshes, and the lack of efficient scaling strategies for higher resolution latent prediction. To address these challenges, we introduce MAR-3D, which integrates a pyramid variational autoencoder with a cascaded masked auto-regressive transformer (Cascaded MAR) for progressive latent upscaling in the continuous space. Our architecture employs random masking during training and auto-regressive denoising in random order during inference, naturally accommodating the unordered property of 3D latent tokens. Additionally, we propose a cascaded training strategy with condition augmentation that enables efficiently up-scale the latent token resolution with fast convergence. Extensive experiments demonstrate that MAR-3D not only achieves superior performance and generalization capabilities compared to existing methods but also exhibits enhanced scaling capabilities compared to joint distribution modeling approaches (e.g., diffusion transformers).

Capturing the temporal evolution of Gaussian properties such as position, rotation, and scale is a challenging task due to the vast number of time-varying parameters and the limited photometric data available, which generally results in convergence issues, making it difficult to find an optimal solution. While feeding all inputs into an end-to- end neural network can effectively model complex temporal dynamics, this approach lacks explicit supervision and struggles to generate high-quality transformation fields. On the other hand, using time-conditioned polynomial functions to model Gaussian trajectories and orientations provides a more explicit and interpretable solution, but requires significant handcrafted effort and lacks generalizability across diverse scenes. To overcome these limitations, this paper introduces a novel approach based on a learnable infinite Taylor Formula to model the temporal evolution of Gaussians. This method offers both the flexibility of an implicit network-based approach and the interpretability of explicit polynomial functions, allowing for more robust and generalizable modeling of Gaussian dynamics across various dynamic scenes. Extensive experiments on dynamic novel view rendering tasks are conducted on public datasets, demonstrating that the proposed method achieves state-of-the-art performance in this domain. More information is available on our project page (https://ellisonking.github.io/TaylorGaussian).

Due to visual ambiguities and inter-person occlusions, existing human pose estimation methods cannot recover plausible close interactions from in-the-wild videos. Even state-of-the-art large foundation models (e.g., SAM) cannot accurately distinguish human semantics in such challenging scenarios. In this work, we find that human appearance can provide a straightforward cue to address these obstacles. Based on this observation, we propose a dual-branch optimization framework to reconstruct accurate interactive motions with plausible body contacts constrained by human appearances, social proxemics, and physical laws. Specifically, we first train a diffusion model to learn the human proxemic behavior and pose prior knowledge. The trained network and two optimizable tensors are then incorporated into a dual-branch optimization framework to reconstruct human motions and appearances. Several constraints based on 3D Gaussians, 2D keypoints, and mesh penetrations are also designed to assist the optimization. With the proxemics prior and diverse constraints, our method is capable of estimating accurate interactions from in-the-wild videos captured in complex environments. We further build a dataset with pseudo ground-truth interaction annotations, which may promote future research on pose estimation and human behavior understanding. Experimental results on several benchmarks demonstrate that our method outperforms existing approaches. The code and data are available at https://www.buzhenhuang.com/works/CloseApp.html.

Reconstructing sharp 3D representations from blurry multi-view images is a long-standing problem in computer vision. Recent works attempt to enhance high-quality novel view synthesis from the motion blur by leveraging event-based cameras, benefiting from high dynamic range and microsecond temporal resolution. However, they often reach sub-optimal visual quality in either restoring inaccurate color or losing fine-grained details. In this paper, we present DiET-GS, a diffusion prior and event stream-assisted motion deblurring 3DGS. Our framework effectively leverages blur-free event streams and diffusion prior in a two-stage training strategy. Specifically, we introduce the novel framework to constrain 3DGS with event double integral, achieving both accurate color and well-defined details. Additionally, we propose a simple technique to leverage diffusion prior to further enhance the edge details. Qualitative and quantitative results on both synthetic and real-world data demonstrate that our DiET-GS is capable of producing better quality of novel views compared to the existing baselines. Our project page is: DiET-GS.github.io.

We introduce TreeMeshGPT, an autoregressive Transformer designed to generate high-quality artistic meshes aligned with input point clouds. Instead of the conventional next-token prediction in autoregressive Transformer, we propose a novel Autoregressive Tree Sequencing where the next input token is retrieved from a dynamically growing tree structure that is built upon the triangle adjacency of faces within the mesh. Our sequencing enables the mesh to extend locally from the last generated triangular face at each step, and therefore reduces training difficulty and improves mesh quality. Our approach represents each triangular face with two tokens, achieving a compression rate of approximately 22% compared to the naive face tokenization. This efficient tokenization enables our model to generate highly detailed artistic meshes with strong point cloud conditioning, surpassing previous methods in both capacity and fidelity. Furthermore, our method generates mesh with strong normal orientation constraints, minimizing flipped normals commonly encountered in previous methods. Our experiments show that TreeMeshGPT enhances the mesh generation quality with refined details and normal orientation consistency.

Identifying affordance regions on 3D objects from semantic cues is essential for robotics and human-machine interaction. However, existing 3D affordance learning methods struggle with generalization and robustness due to limited annotated data and a reliance on 3D backbones focused on geometric encoding, which often lack resilience to real-world noise and data corruption. We propose GEAL, a novel framework designed to enhance the generalization and robustness of 3D affordance learning by leveraging large-scale pre-trained 2D models. We employ a dual-branch architecture with Gaussian splatting to establish consistent mappings between 3D point clouds and 2D representations, enabling realistic 2D renderings from sparse point clouds. A granularity-adaptive fusion module and a 2D-3D consistency alignment module further strengthen cross-modal alignment and knowledge transfer, allowing the 3D branch to benefit from the rich semantics and generalization capacity of 2D models. To holistically assess the robustness, we introduce two new corruption-based benchmarks: PIAD-C and LASO-C. Extensive experiments on public datasets and our benchmarks show that GEAL consistently outperforms existing methods across seen and novel object categories, as well as corrupted data, demonstrating robust and adaptable affordance predictions. The code and datasets are publicly available.

We introduce Generalizable 3D-Language Feature Fields (g3D-LF), a 3D representation model pre-trained on large-scale 3D-language dataset for embodied tasks. Our g3DLF processes posed RGB-D images from agents to encode feature fields for: 1) Novel view representation predictions from any position in the 3D scene; 2) Generations of BEV maps centered on the agent; 3) Querying targets using multi-granularity language within the above-mentioned representations. Our representation can be generalized to unseen environments, enabling real-time construction and online updates. By volume rendering latent features along sampled rays and integrating semantic and spatial relationships through multiscale encoders, our g3D-LF produces representations at different scales and perspectives, aligned with multi-granularity language, via multi-level contrastive learning. Furthermore, we prepare a large-scale 3D-language dataset to align the representations of the feature fields with language. Extensive experiments on Vision-and-Language Navigation under both Panorama and Monocular settings, Zero-shot Object Navigation, and Situated Question Answering tasks highlight the significant advantages and effectiveness of our g3D-LF for embodied tasks. The code is available at https://github.com/MrZihan/g3D-LF.

This work presents IAAO, a novel framework that builds an explicit 3D model for intelligent agents to gain understanding of articulated objects in their environment through interaction. Unlike prior methods that rely on task-specific networks and assumptions about movable parts, our IAAO leverages large foundation models to estimate interactive affordances and part articulations in three stages. We first build hierarchical features and label fields for each object state using 3D Gaussian Splatting (3DGS) by distilling mask features and view-consistent labels from multi-view images. We then perform object- and part-level queries on the 3D Gaussian primitives to identify static and articulated elements, estimating global transformations and local articulation parameters along with affordances. Finally, scenes from different states are merged and refined based on the estimated transformations, enabling robust affordance-based interaction and manipulation of objects. Experimental results demonstrate the effectiveness of our method. Our source code is available at: https://lulusindazc.github.io/IAAOproject/.

Modern 3D semantic scene graph estimation methods utilize ground truth 3D annotations to accurately predict target objects, predicates, and relationships. In the absence of given 3D ground truth representations, we explore leveraging only multi-view RGB images to tackle this task. To attain robust features for accurate scene graph estimation, we must overcome the noisy reconstructed pseudo point-based geometry from predicted depth maps and reduce the amount of background noise present in multi-view image features. The key is to enrich node and edge features with accurate semantic and spatial information and through neighboring relations. We obtain semantic masks to guide feature aggregation to filter background features and design a novel method to incorporate neighboring node information to aid robustness of our scene graph estimates. Furthermore, we leverage on explicit statistical priors calculated from the training summary statistics to refine node and edge predictions based on their one-hop neighborhood. Our experiments show that our method outperforms current methods purely using multi-view images as the initial input. Our project page is available at https://qixun1.github.io/projects/SCRSSG.

Realistic scene reconstruction in driving scenarios poses significant challenges due to fast-moving objects. Most existing methods rely on labor-intensive manual labeling of object poses to reconstruct dynamic objects in canonical space and move them based on these poses during rendering. While some approaches attempt to use 3D object trackers to replace manual annotations, the limited generalization of 3D trackers -- caused by the scarcity of large-scale 3D datasets -- results in inferior reconstructions in real-world settings. In contrast, 2D foundation models demonstrate strong generalization capabilities. To eliminate the reliance on 3D trackers and enhance robustness across diverse environments, we propose a stable object tracking module by leveraging associations from 2D deep trackers within a 3D object fusion strategy. We address inevitable tracking errors by further introducing a motion learning strategy in an implicit feature space that autonomously corrects trajectory errors and recovers missed detections. Experimental results on Waymo-NOTR and KITTI show that our method outperforms existing approaches. Our code will be released on https://lolrudy.github.io/No3DTrackSG/.

High-dynamic scene reconstruction aims to represent static background with rigid spatial features and dynamic objects with deformed continuous spatiotemporal features. Typically, existing methods adopt unified representation model (e.g., Gaussian) to directly match the spatiotemporal features of dynamic scene from frame camera. However, this unified paradigm fails in the potential discontinuous temporal features of objects due to frame imaging and the heterogeneous spatial features between background and objects. To address this issue, we disentangle the spatiotemporal features into various latent representations to alleviate the spatiotemporal mismatching between background and objects. In this work, we introduce event camera to compensate for frame camera, and propose a spatiotemporal-disentangled Gaussian splatting framework for high-dynamic scene reconstruction. As for dynamic scene, we figure out that background and objects have appearance discrepancy in frame-based spatial features and motion discrepancy in event-based temporal features, which motivates us to distinguish the spatiotemporal features between background and objects via clustering. As for dynamic object, we discover that Gaussian representations and event data share the consistent spatiotemporal characteristic, which could serve as a prior to guide the spatiotemporal disentanglement of object Gaussians. Within Gaussian splatting framework, the cumulative scene-object disentanglement can improve the spatiotemporal discrimination between background and objects to render the time-continuous dynamic scene. Extensive experiments have been performed to verify the superiority of the proposed method.

The primary focus of most recent works on open-vocabulary neural fields is extracting precise semantic features from the VLMs and then consolidating them efficiently into a multi-view consistent 3D neural fields representation. However, most existing works over-trusted SAM to regularize image-level CLIP without any further refinement. Moreover, several existing works improved efficiency by dimensionality reduction of semantic features from 2D VLMs before fusing with 3DGS semantic fields, which inevitably leads to multi-view inconsistency. In this work, we propose econSG for open-vocabulary semantic segmentation with 3DGS. Our econSG consists of: 1) A Confidence-region Guided Regularization (CRR) that mutually refines SAM and CLIP to get the best of both worlds for precise semantic features with complete and precise boundaries. 2) A low dimensional contextual space to enforce 3D multi-view consistency while improving computational efficiency by fusing backprojected multi-view 2D features and follow by dimensional reduction directly on the fused 3D features instead of operating on each 2D view separately. Our econSG show state-of-the-art performance on four benchmark datasets compared to the existing methods. Furthermore, we are also the most efficient training among all the methods.

In this work, we tackle the task of learning 3D human Gaussians from a single image, focusing on recovering detailed appearance and geometry including unobserved regions. We introduce a single-view generalizable Human Gaussian Model (HGM), which employs a novel generate-then-refine pipeline with the guidance from human body prior and diffusion prior. Our approach uses a ControlNet to refine rendered back-view images from coarse predicted human Gaussians, then uses the refined image along with the input image to reconstruct refined human Gaussians. To mitigate the potential generation of unrealistic human poses and shapes, we incorporate human priors from the SMPL-X model as a dual branch, propagating image features from the SMPL-X volume to the image Gaussians using sparse convolution and attention mechanisms. Given that the initial SMPL-X estimation might be inaccurate, we gradually refine it with our HGM model. We validate our approach on several publicly available datasets. Our method surpasses previous methods in both novel view synthesis and surface reconstruction. Our approach also exhibits strong generalization for cross-dataset evaluation and in-the-wild images.

Interactive human motions and the continuously changing nature of intentions pose significant challenges for human trajectory prediction. In this paper, we present a neuralized Markov random field (MRF)-based motion evolution method for probabilistic interaction-aware human trajectory prediction. We use MRF to model each agent's motion and the resulting crowd interactions over time, hence is robust against noisy observations and enables group reasoning. We approximate the modeled distribution using two conditional variational autoencoders (CVAEs) for efficient learning and inference. Our proposed method achieves state-of-the-art performance on ADE/FDE metrics across two dataset categories: overhead datasets ETH/UCY, SDD, and NBA, and ego-centric JRDB. Furthermore, our approach allows for real-time stochastic inference in bustling environments, making it well-suited for a 30FPS video setting. We open-source our codes at: https://github.com/AdaCompNUS/NMRF_TrajectoryPrediction.git

3D point clouds directly collected from objects through sensors are often incomplete due to self-occlusion. Conventional methods for completing these partial point clouds rely on manually organized training sets and are usually limited to object categories seen during training. In this work, we propose a test-time framework for completing partial point clouds across unseen categories without any requirement for training. Leveraging point rendering via Gaussian Splatting, we develop techniques of Partial Gaussian Initialization, Zero-shot Fractal Completion, and Point Cloud Extraction that utilize priors from pre-trained 2D diffusion models to infer missing regions and extract uniform completed point clouds. Experimental results on both synthetic and real-world scanned point clouds demonstrate that our approach outperforms existing methods in completing a variety of objects. Our project page is at \url{https://tianxinhuang.github.io/projects/ComPC/}.

In this paper, we investigate Open-Vocabulary 3D Instance Segmentation (OV-3DIS) with free-form language instructions. Earlier works mainly rely on annotated base categories for training which leads to limited generalization to unseen novel categories. To mitigate the poor generalizability to novel categories, recent works generate class-agnostic masks or projecting generalized masks from 2D to 3D, subsequently classifying them with the assistance of 2D foundation model. However, these works often disregard semantic information in the mask generation, leading to sub-optimal performance. Instead, generating generalizable but semantic-aware masks directly from 3D point clouds would result in superior outcomes. To the end, we introduce Segment any 3D Object with LanguagE ($\textbf{SOLE}$), which is a semantic and geometric-aware visual-language learning framework with strong generalizability by generating semantic-related masks directly from 3D point clouds. Specifically, we propose a multimodal fusion network to incorporate multimodal semantics in both backbone and decoder. In addition, to align the 3D segmentation model with various language instructions and enhance the mask quality, we introduce three types of multimodal associations as supervision. Our SOLE outperforms previous methods by a large margin on ScanNetv2, ScanNet200, and Replica benchmarks, and the results are even closed to the fully-supervised counterpart despite the absence of class annotations in the training. Furthermore, extensive qualitative results demonstrate the versatility of our SOLE to language instructions. The code will be made publicly available.

Recent developments in 2D visual generation have been remarkably successful. However, 3D and 4D generation remain challenging in real-world applications due to the lack of large-scale 4D data and effective model design. In this paper, we propose to jointly investigate general 3D and 4D generation by leveraging camera and object movements commonly observed in daily life. Due to the lack of real-world 4D data in the community, we first propose a data curation pipeline to obtain camera poses and object motion strength from videos. Based on this pipeline, we introduce a large-scale real-world 4D scene dataset: CamVid-30K. By leveraging all the 3D and 4D data, we develop our framework, GenXD, which allows us to produce any 3D or 4D scene. We propose multiview-temporal modules, which disentangle camera and object movements, to seamlessly learn from both 3D and 4D data. Additionally, GenXD employs masked latent conditions to support a variety of conditioning views. GenXD can generate videos that follow the camera trajectory as well as consistent 3D views that can be lifted into 3D representations. We perform extensive evaluations across various real-world and synthetic datasets, demonstrating GenXD's effectiveness and versatility compared to previous methods in 3D and 4D generation.

This paper proposes a versatile graph-based lifelong localization framework using LiDAR, LiLoc, which enhances its timeliness by maintaining a single central session while improves the accuracy through multi-modal factors between the central and subsidiary sessions. First, an adaptive submap joining strategy is employed to generate prior submaps (keyframes and poses) for the central session, and to provide priors for subsidiaries when constraints are needed for robust localization. Next, a coarse-to-fine pose initialization for subsidiary sessions is performed using vertical recognition and ICP refinement in the global coordinate frame. To elevate the accuracy of subsequent localization, we propose an egocentric factor graph (EFG) module that integrates the IMU preintegration, LiDAR odometry and scan match factors in a joint optimization manner. Specifically, the scan match factors are constructed by a novel propagation model that efficiently distributes the prior constrains as edges to the relevant prior pose nodes, weighted by noises based on keyframe registration errors. Additionally, the framework supports flexible switching between two modes: relocalization (RLM) and incremental localization (ILM) based on the proposed overlap-based mechanism to select or update the prior submaps from central session. The proposed LiLoc is tested on public and custom datasets, demonstrating accurate localization performance against state-of-the-art methods. Our codes will be publicly available on https://github.com/Yixin-F/LiLoc.

Neural radiance fields (NeRF) and 3D Gaussian Splatting (3DGS) are popular techniques to reconstruct and render photorealistic images. However, the prerequisite of running Structure-from-Motion (SfM) to get camera poses limits their completeness. Although previous methods can reconstruct a few unposed images, they are not applicable when images are unordered or densely captured. In this work, we propose a method to train 3DGS from unposed images. Our method leverages a pre-trained 3D geometric foundation model as the neural scene representation. Since the accuracy of the predicted pointmaps does not suffice for accurate image registration and high-fidelity image rendering, we propose to mitigate the issue by initializing and fine-tuning the pre-trained model from a seed image. The images are then progressively registered and added to the training buffer, which is used to train the model further. We also propose to refine the camera poses and pointmaps by minimizing a point-to-camera ray consistency loss across multiple views. When evaluated on diverse challenging datasets, our method outperforms state-of-the-art pose-free NeRF/3DGS methods in terms of both camera pose accuracy and novel view synthesis, and even renders higher fidelity images than 3DGS trained with COLMAP poses.

Novel view synthesis from monocular videos of dynamic scenes with unknown camera poses remains a fundamental challenge in computer vision and graphics. While recent advances in 3D representations such as Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS) have shown promising results for static scenes, they struggle with dynamic content and typically rely on pre-computed camera poses. We present 4D3R, a pose-free dynamic neural rendering framework that decouples static and dynamic components through a two-stage approach. Our method first leverages 3D foundational models for initial pose and geometry estimation, followed by motion-aware refinement. 4D3R introduces two key technical innovations: (1) a motion-aware bundle adjustment (MA-BA) module that combines transformer-based learned priors with SAM2 for robust dynamic object segmentation, enabling more accurate camera pose refinement; and (2) an efficient Motion-Aware Gaussian Splatting (MA-GS) representation that uses control points with a deformation field MLP and linear blend skinning to model dynamic motion, significantly reducing computational cost while maintaining high-quality reconstruction. Extensive experiments on real-world dynamic datasets demonstrate that our approach achieves up to 1.8dB PSNR improvement over state-of-the-art methods, particularly in challenging scenarios with large dynamic objects, while reducing computational requirements by 5× compared to previous dynamic scene representations.

Event cameras are neuromorphic vision sensors that asynchronously capture pixel-level intensity changes with high temporal resolution and dynamic range. These make them well suited for monocular depth estimation under challenging lighting conditions. However, progress in event-based monocular depth estimation remains constrained by the quality of supervision: LiDAR-based depth labels are inherently sparse, spatially incomplete, and prone to artifacts. Consequently, these signals are suboptimal for learning dense depth from sparse events. To address this problem, we propose Distil-E2D, a framework that distills depth priors from the image domain into the event domain by generating dense synthetic pseudolabels from co-recorded APS or RGB frames using foundational depth models. These pseudolabels complement sparse LiDAR depths with dense semantically rich supervision informed by large-scale image-depth datasets. To reconcile discrepancies between synthetic and real depths, we introduce a Confidence-Guided Calibrated Depth Loss that learns nonlinear depth alignment and adaptively weights supervision by alignment confidence. Additionally, our architecture integrates past predictions via a Context Transformer and employs a Dual-Decoder Training scheme that enhances encoder representations by jointly learning metric and relative depth abstractions. Experiments on benchmark datasets show that Distil-E2D achieves state-of-the-art performance in event-based monocular depth estimation across both event-only and event+APS settings.

Event cameras offer unparalleled advantages for real-time perception in dynamic environments, thanks to the microsecond-level temporal resolution and asynchronous operation. Existing event detectors, however, are limited by fixed-frequency paradigms and fail to fully exploit the high-temporal resolution and adaptability of event data. To address these limitations, we propose FlexEvent, a novel framework that enables detection at varying frequencies. Our approach consists of two key components: FlexFuse, an adaptive event-frame fusion module that integrates high-frequency event data with rich semantic information from RGB frames, and FlexTune, a frequency-adaptive fine-tuning mechanism that generates frequency-adjusted labels to enhance model generalization across varying operational frequencies. This combination allows our method to detect objects with high accuracy in both fast-moving and static scenarios, while adapting to dynamic environments. Extensive experiments on large-scale event camera datasets demonstrate that our approach surpasses state-of-the-art methods, achieving significant improvements in both standard and high-frequency settings. Notably, our method maintains robust performance when scaling from 20 Hz to 90 Hz and delivers accurate detection up to 180 Hz, proving its effectiveness in extreme conditions. Our framework sets a new benchmark for event-based object detection and paves the way for more adaptable, real-time vision systems.

Pairwise camera pose estimation from sparsely overlapping image pairs remains a critical and unsolved challenge in 3D vision. Most existing methods struggle with image pairs that have small or no overlap. Recent approaches attempt to address this by synthesizing intermediate frames using video interpolation and selecting key frames via a self-consistency score. However, the generated frames are often blurry due to small overlap inputs, and the selection strategies are slow and not explicitly aligned with pose estimation. To solve these cases, we propose Hybrid Video Generation (HVG) to synthesize clearer intermediate frames by coupling a video interpolation model with a pose-conditioned novel view synthesis model, where we also propose a Feature Matching Selector (FMS) based on feature correspondence to select intermediate frames appropriate for pose estimation from the synthesized results. Extensive experiments on Cambridge Landmarks, ScanNet, DL3DV-10K, and NAVI demonstrate that, compared to existing SOTA methods, PoseCrafter can obviously enhance the pose estimation performances, especially on examples with small or no overlap.

Vision-and-Language Navigation (VLN) is a core task where embodied agents leverage their spatial mobility to navigate in 3D environments toward designated destinations based on natural language instructions. Recently, video-language large models (Video-VLMs) with strong generalization capabilities and rich commonsense knowledge have shown remarkable performance when applied to VLN tasks. However, these models still encounter the following challenges when applied to real-world 3D navigation: 1) Insufficient understanding of 3D geometry and spatial semantics; 2) Limited capacity for large-scale exploration and long-term environmental memory; 3) Poor adaptability to dynamic and changing environments.To address these limitations, we propose Dynam3D, a dynamic layered 3D representation model that leverages language-aligned, generalizable, and hierarchical 3D representations as visual input to train 3D-VLM in navigation action prediction. Given posed RGB-D images, our Dynam3D projects 2D CLIP features into 3D space and constructs multi-level 3D patch-instance-zone representations for 3D geometric and semantic understanding with a dynamic and layer-wise update strategy. Our Dynam3D is capable of online encoding and localization of 3D instances, and dynamically updates them in changing environments to provide large-scale exploration and long-term memory capabilities for navigation. By leveraging large-scale 3D-language pretraining and task-specific adaptation, our Dynam3D sets new state-of-the-art performance on VLN benchmarks including R2R-CE, REVERIE-CE and NavRAG-CE under monocular settings. Furthermore, experiments for pre-exploration, lifelong memory, and real-world robot validate the effectiveness of practical deployment.

Depth completion in dynamic scenes poses significant challenges due to rapid ego-motion and object motion, which can severely degrade the quality of input modalities such as RGB images and LiDAR measurements. Conventional RGB-D sensors often struggle to align precisely and capture reliable depth under such conditions. In contrast, event cameras with their high temporal resolution and sensitivity to motion at the pixel level provide complementary cues that are beneficial in dynamic environments. To this end, we propose EventDC, the first event-driven depth completion framework. It consists of two key components: Event-Modulated Alignment (EMA) and Local Depth Filtering (LDF). Both modules adaptively learn the two fundamental components of convolution operations: offsets and weights conditioned on motion-sensitive event streams. In the encoder, EMA leverages events to modulate the sampling positions of RGB-D features to achieve pixel redistribution for improved alignment and fusion. In the decoder, LDF refines depth estimations around moving objects by learning motion-aware masks from events. Additionally, EventDC incorporates two loss terms to further benefit global alignment and enhance local depth recovery. Moreover, we establish the first benchmark for event-based depth completion comprising one real-world and two synthetic datasets to facilitate future research. Extensive experiments on this benchmark demonstrate the superiority of our EventDC. [Project page](https://yanzq95.github.io/projectpage/EventDC/index.html).

Diffusion models are advancing autonomous driving by enabling realistic data synthesis, predictive end-to-end planning, and closed-loop simulation, with a primary focus on temporally consistent generation. However, large-scale 3D scene generation requiring spatial coherence remains underexplored. In this paper, we present X-Scene, a novel framework for large-scale driving scene generation that achieves geometric intricacy, appearance fidelity, and flexible controllability. Specifically, X-Scene supports multi-granular control, including low-level layout conditioning driven by user input or text for detailed scene composition, and high-level semantic guidance informed by user intent and LLM-enriched prompts for efficient customization. To enhance geometric and visual fidelity, we introduce a unified pipeline that sequentially generates 3D semantic occupancy and corresponding multi-view images and videos, ensuring alignment and temporal consistency across modalities. We further extend local regions into large-scale scenes via consistency-aware outpainting, which extrapolates occupancy and images from previously generated areas to maintain spatial and visual coherence. The resulting scenes are lifted into high-quality 3DGS representations, supporting diverse applications such as simulation and scene exploration. Extensive experiments demonstrate that X-Scene substantially advances controllability and fidelity in large-scale scene generation, empowering data generation and simulation for autonomous driving.

Recent advancements in foundation models for 2D vision have substantially improved the analysis of dynamic scenes from monocular videos. However, despite their strong generalization capabilities, these models often lack 3D consistency, a fundamental requirement for understanding scene geometry and motion, thereby causing severe spatial misalignment and temporal flickering in complex 3D environments. In this paper, we present Motion4D, a novel framework that addresses these challenges by integrating 2D priors from foundation models into a unified 4D Gaussian Splatting representation. Our method features a two-part iterative optimization framework: 1) Sequential optimization, which updates motion and semantic fields in consecutive stages to maintain local consistency, and 2) Global optimization, which jointly refines all attributes for long-term coherence. To enhance motion accuracy, we introduce a 3D confidence map that dynamically adjusts the motion priors, and an adaptive resampling process that inserts new Gaussians into under-represented regions based on per-pixel RGB and semantic errors. Furthermore, we enhance semantic coherence through an iterative refinement process that resolves semantic inconsistencies by alternately optimizing the semantic fields and updating prompts of SAM2. Extensive evaluations demonstrate that our Motion4D significantly outperforms both 2D foundation models and existing 3D-based approaches across diverse scene understanding tasks, including point-based tracking, video object segmentation, and novel view synthesis. Our code is available at https://hrzhou2.github.io/motion4d-web/.

We introduce DiHuR, a novel Diffusion-guided model for generalizable Human 3D Reconstruction and view synthe-sis from sparse, minimally overlapping images. While existing generalizable human radiance fields excel at novel view synthesis, they often struggle with comprehensive 3D reconstruction. Similarly, directly optimizing implicit Signed Distance Function (SDF) fields from sparse-view images typically yields poor results due to limited overlap. To enhance 3D reconstruction quality, we propose using learnable tokens associated with SMPL vertices to aggregate sparse view features and then to guide SDF prediction. These tokens learn a generalizable prior across different identities in training datasets, leveraging the consistent projection of SMPL vertices onto similar semantic areas across various human identities. This consistency enables effective knowledge transfer to unseen identities during inference. Recognizing SMPL's limitations in capturing clothing details, we incorporate a diffusion model as an additional prior to fill in missing information, particularly for complex clothing geometries. Our method integrates two key priors in a coherent manner: the prior from generaliz-able feed-forward models and the 2D diffusion prior, and it requires only multi-view image training, without 3D supervision. DiHuR demonstrates superior performance in both within-dataset and cross-dataset generalization settings, as validated on THuman, ZJU-MoCap, and HuMMan datasets compared to existing methods.

Deep learning approaches have demonstrated high effectiveness in 3D object detection tasks. However, they often suffer from a notable drop in performance on the previously trained classes when learning new classes incrementally without revisiting the old data. This is the "catastrophic forgetting" phenomenon which impedes 3D object detection in real-world scenarios, where intelligent machines must continuously learn to detect previously unseen categories. Furthermore, frequent co-occurrences of old and new classes in scenes exacerbate catastrophic forgetting and cause model confusion. To address these challenges, we propose a novel static-dynamic co-teaching approach. Our framework involves a student model and two teacher models: a static teacher with fixed weights which imparts preserved old knowledge to the student, and a dynamic teacher with continuously updated weights which transfers underlying knowledge from new data to the student. To mitigate the issue of co-occurrence, we generate pseudo labels for base (i.e. old) classes from both static and dynamic sources during incremental learning. Additionally, to mitigate the negative impact of varying occurrence frequencies of classes on fixed thresholding during the selection of pseudo labels, we calibrate the probabilities of base classes to attain more balanced class probabilities. Moreover, our static-dynamic co-teaching framework is backbone-agnostic, making it compatible with different detection architectures. We demonstrate its backbone-agnostic nature by adapting three representative 3D object detectors: VoteNet, 3DETR and CAGroup3D. Extensive experiments showcase the superior performance of our proposed method compared to baseline approaches across indoor and outdoor benchmark datasets and applicability with different backbone models.

In this paper, we tackle the challenging task of learning a generalizable human NeRF model from a monocular video. Although existing generalizable human NeRFs have achieved impressive results, they require muti-view images or videos which might not be always available. On the other hand, some works on free-viewpoint rendering of human from monocular videos cannot be generalized to unseen identities. In view of these limitations, we propose GHuNeRF to learn a generalizable human NeRF model from a monocular video of the human performer. We first introduce a visibility-aware aggregation scheme to compute vertex-wise features, which is used to construct a 3D feature volume. The feature volume can only represent the overall geometry of the human performer with insufficient accuracy due to the limited resolution. To solve this, we further enhance the volume feature with temporally aligned pointwise features using an attention mechanism. Finally, the enhanced feature is used for predicting density and color for each sampled point. A surface-guided sampling strategy is also adopted to improve the efficiency for both training and inference. We validate our approach on the widely-used ZJU-MoCap dataset, where we achieve comparable performance with existing multi-view video based approaches. We also test on the monocular People-Snapshot dataset and achieve better performance than existing works when only monocular video is used. Our code is available at the project website1.1https://github.com/chaneyddtt/GHuNeRF

The field of self-supervised $3 D$ representation learning has emerged as a promising solution to alleviate the challenge presented by the scarcity of extensive, well-annotated datasets. However, it continues to be hindered by the lack of diverse, large-scale, real-world 3D scene datasets for source data. To address this shortfall, we propose Generalizable Representation Learning (GRL), where we devise a generative Bayesian network to produce diverse synthetic scenes with real-world patterns, and conduct pre-training with a joint objective. By jointly learning a coarse-to-fine contrastive learning task and an occlusion-aware reconstruction task, the model is primed with transferable, geometry-informed representations. Post pre-training on synthetic data, the acquired knowledge of the model can be seamlessly transferred to two principal downstream tasks associated with 3D scene understanding, namely 3D object detection and 3D semantic segmentation, using real-world benchmark datasets. A thorough series of experiments robustly display our method’s consistent superiority over existing state-of-the-art pre-training approaches.

Learning from open-world noisy data, where both closed-set and open-set noise co-exist in the dataset, is a realistic but underexplored setting. Only recently, several efforts have been initialized to tackle this problem. However, these works assume the classes are balanced when dealing with open-world noisy data. This assumption often violates the nature of real-world large-scale datasets, where the label distributions are generally long-tailed, i.e. class-imbalanced. In this paper, we study the problem of robust visual recognition with class-imbalanced open-world noisy data. We propose a probabilistic graphical model-based approach: iMRF to achieve label noise correction that is robust to class imbalance via an efficient iterative inference of a Markov Random Field (MRF) in each training mini-batch. Furthermore, we design an agreement-based thresholding strategy to adaptively collect clean samples from all classes that includes corrected closed-set noisy samples while rejecting open-set noisy samples. We also introduce a noise-aware balanced cross-entropy loss to explicitly eliminate the bias caused by class-imbalanced data. Extensive experiments on several benchmark datasets including synthetic and real-world noisy datasets demonstrate the superior performance robustness of our method over existing methods. Our code is available at https://github.com/Na-Z/LIOND.

Existing multi-person human reconstruction approaches mainly focus on recovering accurate poses or avoiding penetration, but overlook the modeling of close interactions. In this work, we tackle the task of reconstructing closely interactive humans from a monocular video. The main challenge of this task comes from insufficient visual information caused by depth ambiguity and severe inter-person occlusion. In view of this, we propose to leverage knowledge from proxemic behavior and physics to compensate the lack of visual information. This is based on the observation that human interaction has specific patterns following the social proxemics. Specifically, we first design a latent representation based on Vector Quantised-Variational AutoEncoder (VQ-VAE) to model human interaction. A proxemics and physics guided diffusion model is then introduced to denoise the initial distribution. We design the diffusion model as dual branch with each branch representing one individual such that the interaction can be modeled via cross attention. With the learned priors of VQ-VAE and physical constraint as the additional information, our proposed approach is capable of estimating accurate poses that are also proxemics and physics plausible. Experimental results on Hi4D, 3DPW, and CHI3D demonstrate that our method outperforms existing approaches. The code is available at https://github.com/boycehbz/HumanInteraction.

We present DiSR-NeRF, a diffusion-guided framework for view-consistent super-resolution (SR) NeRF. Unlike prior works, we circumvent the requirement for high-resolution (HR) reference images by leveraging existing powerful 2D super-resolution models. Nonetheless, independent SR 2D images are often inconsistent across different views. We thus propose Iterative 3D Synchronization (I3DS) to mitigate the inconsistency problem via the inherent multi-view consistency property of NeRF. Specifically, our I3DS alternates between upscaling low-resolution (LR) rendered images with diffusion models, and updating the underlying 3D representation with standard NeRF training. We further introduce Renoised Score Distillation (RSD), a novel score-distillation objective for 2D image resolution. Our RSD combines features from ancestral sampling and Score Distillation Sampling (SDS) to generate sharp images that are also LR-consistent. Qualitative and quantitative results on both synthetic and real-world datasets demonstrate that our DiSR-NeRF can achieve better results on NeRF super-resolution compared with existing works. Code and video results available at the project website’,

In this paper, we tackle the task of category-agnostic pose estimation (CAPE), which aims to predict poses for objects of any category with few annotated samples. Previous works either rely on local matching between features of support and query samples or require support keypoint identifier. The former is prone to overfitting due to its sensitivity to sparse samples, while the latter is impractical for the open-world nature of the task. To overcome these limitations, we propose ESCAPE - a Bayesian framework that learns a prior over the features of keypoints. The prior can be expressed as a mixture of super-keypoints, each being a high-level abstract keypoint that captures the statistics of semantically related keypoints from different categories. We estimate the super-keypoints from base categories and use them in adaptation to novel categories. The adaptation to an unseen category involves two steps: first, we match each novel keypoint to a related superkeypoint; and second, we transfer the knowledge encoded in the matched super-keypoints to the novel keypoints. For the first step, we propose a learnable matching network to capture the relationship between the novel keypoints and the super-keypoints, resulting in a more reliable matching. ESCAPE mitigates overfitting by directly transferring learned knowledge to novel categories while it does not use keypoint identifiers. We achieve state-of-the-art performance on the standard MP-100 benchmark. Our code is available at https://github.com/khoiucd/escape-tgt.

Recent advancements in vision-language foundation models have significantly enhanced open-vocabulary 3D scene understanding. However, the generalizability of existing methods is constrained due to their framework designs and their reliance on 3D data. We address this limitation by introducing Generalizable Open-Vocabulary Neural Semantic Fields (GOV-NeSF), a novel approach offering a generalizable implicit representation of 3D scenes with open-vocabulary semantics. We aggregate the geometry-aware features using a cost volume, and propose a Multi-view Joint Fusion module to aggregate multi-view features through a cross-view attention mechanism, which effectively predicts view-specific blending weights for both colors and open-vocabulary features. Remarkably, our GOV-NeSF exhibits state-of-the-art performance in both 2D and 3D open-vocabulary semantic segmentation, eliminating the need for ground truth semantic labels or depth priors, and effectively generalize across scenes and datasets without fine-tuning.

3D Gaussians have recently emerged as a highly efficient representation for 3D reconstruction and rendering. Despite its high rendering quality and speed at high resolutions, they both deteriorate drastically when rendered at lower resolutions or from far away camera position. During low resolution or far away rendering, the pixel size of the image can fall below the Nyquist frequency compared to the screen size of each splatted 3D Gaussian and leads to aliasing effect. The rendering is also drastically slowed down by the sequential alpha blending of more splatted Gaussians per pixel. To address these issues, we propose a multi-scale 3D Gaussian splatting algorithm, which maintains Gaussians at different scales to represent the same scene. Higher-resolution images are rendered with more small Gaussians, and lower-resolution images are rendered with fewer larger Gaussians. With similar training time, our algorithm can achieve 13%-66% PSNR and 160%-2400% rendering speed improvement at 4 × -128 × scale rendering on Mip-NeRF360 dataset compared to the single scale 3D Gaussian splatting. More results and code are released on our project page.

We introduce X-Ray, a novel 3D sequential representation inspired by the penetrability of x-ray scans. X-Ray transforms a 3D object into a series of surface frames at different layers, making it suitable for generating 3D models from images. Our method utilizes ray casting from the camera center to capture geometric and textured details, including depth, normal, and color, across all intersected surfaces. This process efficiently condenses the whole 3D object into a multi-frame video format, motivating the utilize of a network architecture similar to those in video diffusion models. This design ensures an efficient 3D representation by focusing solely on surface information. Also, we propose a two-stage pipeline to generate 3D objects from X-Ray Diffusion Model and Upsampler. We demonstrate the practicality and adaptability of our X-Ray representation by synthesizing the complete visible and hidden surfaces of a 3D object from a single input image. Experimental results reveal the state-of-the-art superiority of our representation in enhancing the accuracy of 3D generation, paving the way for new 3D representation research and practical applications. Our project page is in \url{https://tau-yihouxiang.github.io/projects/X-Ray/X-Ray.html}.

The recent advances in 3D Gaussian Splatting (3DGS) show promising results on the novel view synthesis (NVS) task. With its superior rendering performance and high-fidelity rendering quality, 3DGS is excelling at its previous NeRF counterparts. The most recent 3DGS method focuses either on improving the instability of rendering efficiency or reducing the model size. On the other hand, the training efficiency of 3DGS on large-scale scenes has not gained much attention. In this work, we propose DoGaussian, a method that trains 3DGS distributedly. Our method first decomposes a scene into $K$ blocks and then introduces the Alternating Direction Method of Multipliers (ADMM) into the training procedure of 3DGS. During training, our DoGaussian maintains one global 3DGS model on the master node and $K$ local 3DGS models on the slave nodes. The $K$ local 3DGS models are dropped after training and we only query the global 3DGS model during inference. The training time is reduced by scene decomposition, and the training convergence and stability are guaranteed through the consensus on the shared 3D Gaussians. Our method accelerates the training of 3DGS by $6+$ times when evaluated on large-scale scenes while concurrently achieving state-of-the-art rendering quality. Our code is publicly available at [https://github.com/AIBluefisher/DOGS](https://github.com/AIBluefisher/DOGS).

Although 3D Gaussian Splatting has been widely studied because of its realistic and efficient novel-view synthesis, it is still challenging to extract a high-quality surface from the point-based representation. Previous works improve the surface by incorporating geometric priors from the off-the-shelf normal estimator. However, there are two main limitations: 1) Supervising normal rendered from 3D Gaussians updates only the rotation parameter while neglecting other geometric parameters; 2) The inconsistency of predicted normal maps across multiple views may lead to severe reconstruction artifacts. In this paper, we propose a Depth-Normal regularizer that directly couples normal with other geometric parameters, leading to full updates of the geometric parameters from normal regularization. We further propose a confidence term to mitigate inconsistencies of normal predictions across multiple views. Moreover, we also introduce a densification and splitting strategy to regularize the size and distribution of 3D Gaussians for more accurate surface modeling. Compared with Gaussian-based baselines, experiments show that our approach obtains better reconstruction quality and maintains competitive appearance quality at faster training speed and 100+ FPS rendering. Our code will be made open-source upon paper acceptance.

Existing research has made impressive strides in reconstructing human facial shapes and textures from images with well-illuminated faces and minimal external occlusions. Nevertheless, it remains challenging to recover accurate facial textures from scenarios with complicated illumination affected by external occlusions, \eg a face that is partially obscured by items such as a hat. Existing works based on the assumption of single and uniform illumination cannot correctly process these data. In this work, we introduce a novel approach to model 3D facial textures under such unnatural illumination. Instead of assuming single illumination, our framework learns to imitate the unnatural illumination as a composition of multiple separate light conditions combined with learned neural representations, named Light Decoupling. According to experiments on both single images and video sequences, we demonstrate the effectiveness of our approach in modeling facial textures under challenging illumination affected by occlusions.

Empowering 3D Gaussian Splatting with generalization ability is appealing. However, existing generalizable 3D Gaussian Splatting methods are largely confined to narrow-range interpolation between stereo images due to their heavy backbones, thus lacking the ability to accurately localize 3D Gaussian and support free-view synthesis across wide view range. In this paper, we present a novel framework FreeSplat that is capable of reconstructing geometrically consistent 3D scenes from long sequence input towards free-view synthesis.Specifically, we firstly introduce Low-cost Cross-View Aggregation achieved by constructing adaptive cost volumes among nearby views and aggregating features using a multi-scale structure. Subsequently, we present the Pixel-wise Triplet Fusion to eliminate redundancy of 3D Gaussians in overlapping view regions and to aggregate features observed across multiple views. Additionally, we propose a simple but effective free-view training strategy that ensures robust view synthesis across broader view range regardless of the number of views. Our empirical results demonstrate state-of-the-art novel view synthesis peformances in both novel view rendered color maps quality and depth maps accuracy across different numbers of input views. We also show that FreeSplat performs inference more efficiently and can effectively reduce redundant Gaussians, offering the possibility of feed-forward large scene reconstruction without depth priors. Our code will be made open-source upon paper acceptance.

The key challenge of multi-view indoor 3D object detection is to infer accurate geometry information from images for precise 3D detection. Previous method relies on NeRF for geometry reasoning. However, the geometry extracted from NeRF is generally inaccurate, which leads to sub-optimal detection performance. In this paper, we propose MVSDet which utilizes plane sweep for geometry-aware 3D object detection. To circumvent the requirement for a large number of depth planes for accurate depth prediction, we design a probabilistic sampling and soft weighting mechanism to decide the placement of pixel features on the 3D volume. We select multiple locations that score top in the probability volume for each pixel and use their probability score to indicate the confidence. We further apply recent pixel-aligned Gaussian Splatting to regularize depth prediction and improve detection performance with little computation overhead. Extensive experiments on ScanNet and ARKitScenes datasets are conducted to show the superiority of our model. Our code is available at https://github.com/Pixie8888/MVSDet.

Existing works on weakly-supervised audio-visual video parsing adopt hybrid attention network (HAN) as the multi-modal embedding to capture the cross-modal context. It embeds the audio and visual modalities with a shared network, where the cross-attention is performed at the input. However, such an early fusion method highly entangles the two non-fully correlated modalities and leads to sub-optimal performance in detecting single-modality events. To deal with this problem, we propose the messenger-guided mid-fusion transformer to reduce the uncorrelated cross-modal context in the fusion. The messengers condense the full cross-modal context into a compact representation to only preserve useful cross-modal information. Furthermore, due to the fact that microphones capture audio events from all directions, while cameras only record visual events within a restricted field of view, there is a more frequent occurrence of unaligned cross-modal context from audio for visual event predictions. We thus propose cross-audio prediction consistency to suppress the impact of irrelevant audio information on visual event prediction. Experiments consistently illustrate the superior performance of our framework compared to existing state-of-the-art methods.

Incremental few-shot object detection aims at detecting novel classes without forgetting knowledge of the base classes with only a few labeled training data from the novel classes. Most related prior works are on incremental object detection that rely on the availability of abundant training samples per novel class that substantially limits the scalability to real-world setting where novel data can be scarce. In this paper, we propose the Incremental-DETR that does incremental few-shot object detection via fine-tuning and self-supervised learning on the DETR object detector. To alleviate severe over-fitting with few novel class data, we first fine-tune the class-specific components of DETR with self-supervision from additional object proposals generated using Selective Search as pseudo labels. We further introduce an incremental few-shot fine-tuning strategy with knowledge distillation on the class-specific components of DETR to encourage the network in detecting novel classes without forgetting the base classes. Extensive experiments conducted on standard incremental object detection and incremental few-shot object detection settings show that our approach significantly outperforms state-of-the-art methods by a large margin. Our source code is available at https://github.com/dongnana777/Incremental-DETR.

Animal pose estimation is an important but underexplored task due to the lack of labeled data. In this paper, we tackle the task of animal pose estimation with scarce annotations, where only a small set of labeled data and unlabeled images are available. At the core of the solution to this problem setting is the use of the unlabeled data to compensate for the lack of well-labeled animal pose data. To this end, we propose the ScarceNet, a pseudo label-based approach to generate artificial labels for the unlabeled images. The pseudo labels, which are generated with a model trained with the small set of labeled images, are generally noisy and can hurt the performance when directly used for training. To solve this problem, we first use a small-loss trick to select reliable pseudo labels. Although effective, the selection process is improvident since numerous high-loss samples are left unused. We further propose to identify reusable samples from the high-loss samples based on an agreement check. Pseudo labels are re-generated to provide supervision for those reusable samples. Lastly, we introduce a student-teacher framework to enforce a consistency constraint since there are still samples that are neither reliable nor reusable. By combining the reliable pseudo label selection with the reusable sample re-labeling and the consistency constraint, we can make full use of the unlabeled data. We evaluate our approach on the challenging AP-10K dataset, where our approach outperforms existing semi-supervised approaches by a large margin. We also test on the TigDog dataset, where our approach can achieve better performance than domain adaptation based approaches when only very few annotations are available. Our code is available at the project website11https://github.com/chaneyddtt/ScarceNet.

Recent works such as BARF and GARF can bundle adjust camera poses with neural radiance fields (NeRF) which is based on coordinate-MLPs. Despite the impressive results, these methods cannot be applied to Generalizable NeRFs (GeNeRFs) which require image feature extractions that are often based on more complicated 3D CNN or transformer architectures. In this work, we first analyze the difficulties of jointly optimizing camera poses with GeNeRFs, and then further propose our DBARF to tackle these issues. Our DBARF which bundle adjusts camera poses by taking a cost feature map as an implicit cost function can be jointly trained with GeNeRFs in a self-supervised manner. Unlike BARF and its follow-up works, which can only be applied to per-scene optimized NeRFs and need accurate initial camera poses with the exception of forward-facing scenes, our method can generalize across scenes and does not require any good initialization. Experiments show the effectiveness and generalization ability of our DBARF when evaluated on real-world datasets. Our code is available at https://aibluefisher.github.io/DBARF.

Dynamic Neural Radiance Field (NeRF) is a powerful algorithm capable of rendering photo-realistic novel view images from a monocular RGB video of a dynamic scene. Although it warps moving points across frames from the observation spaces to a common canonical space for rendering, dynamic NeRF does not model the change of the reflected color during the warping. As a result, this approach often fails drastically on challenging specular objects in motion. We address this limitation by reformulating the neural radiance field function to be conditioned on surface position and orientation in the observation space. This allows the specular surface at different poses to keep the different reflected colors when mapped to the common canonical space. Additionally, we add the mask of moving objects to guide the deformation field. As the specular surface changes color during motion, the mask mitigates the problem of failure to find temporal correspondences with only RGB supervision. We evaluate our model based on the novel view synthesis quality with a self-collected dataset of different moving specular objects in realistic environments. The experimental results demonstrate that our method significantly improves the reconstruction quality of moving specular objects from monocular RGB videos compared to the existing NeRF models. Our code and data are available at the project website11https://github.com/JokerYan/NeRF-DS.

Although many recent works have investigated generalizable NeRF-based novel view synthesis for unseen scenes, they seldom consider the synthetic-to-real generalization, which is desired in many practical applications. In this work, we first investigate the effects of synthetic data in synthetic-to-real novel view synthesis and surprisingly observe that models trained with synthetic data tend to produce sharper but less accurate volume densities. For pixels where the volume densities are correct, fine-grained details will be obtained. Otherwise, severe artifacts will be produced. To maintain the advantages of using synthetic data while avoiding its negative effects, we propose to introduce geometry-aware contrastive learning to learn multi-view consistent features with geometric constraints. Meanwhile, we adopt cross-view attention to further enhance the geometry perception of features by querying features across input views. Experiments demonstrate that under the synthetic-to-real setting, our method can render images with higher quality and better fine-grained details, outperforming existing generalizable novel view synthesis methods in terms of PSNR, SSIM, and LPIPS. When trained on real data, our method also achieves state-of-the-art results. https://haoy945.github.io/contranerf/

Direct mesh fitting for 3D hand shape reconstruction is highly accurate. However, the reconstructed meshes are prone to artifacts and do not appear as plausible hand shapes. Conversely, parametric models like MANO ensure plausible hand shapes but are not as accurate as the non-parametric methods. In this work, we introduce a novel weakly-supervised hand shape estimation framework that integrates non-parametric mesh fitting with MANO model in an end-to-end fashion. Our joint model overcomes the tradeoff in accuracy and plausibility to yield well-aligned and high-quality 3D meshes, especially in challenging two-hand and hand-object interaction scenarios.

The main challenges of 3D pose transfer are: 1) Lack of paired training data with different characters performing the same pose; 2) Disentangling pose and shape information from the target mesh; 3) Difficulty in applying to meshes with different topologies. We thus propose a novel weakly-supervised keypoint-based framework to overcome these difficulties. Specifically, we use a topology-agnostic keypoint detector with inverse kinematics to compute transformations between the source and target meshes. Our method only requires supervision on the keypoints, can be applied to meshes with different topologies and is shape-invariant for the target which allows extraction of pose-only information from the target meshes without transferring shape information. We further design a cycle reconstruction to perform self-supervised pose transfer without the need for ground truth deformed mesh with the same pose and shape as the target and source, respectively. We evaluate our approach on benchmark human and animal datasets, where we achieve superior performance compared to the state-of-the-art unsupervised approaches and even comparable performance with the fully supervised approaches. We test on the more challenging Mixamo dataset to verify our approach’s ability in handling meshes with different topologies and complex clothes. Cross-dataset evaluation further shows the strong generalization ability of our approach. Our source code is available at: https://github.com/jinnan-chen/3D-Pose-Transfer.

Although Neural Radiance Fields (NeRF) is popular in the computer vision community recently, registering multiple NeRFs has yet to gain much attention. Unlike the existing work, NeRF2NeRF [14], which is based on traditional optimization methods and needs human annotated keypoints, we propose DReg-NeRF to solve the NeRF registration problem on object-centric scenes without human intervention. After training NeRF models, our DReg-NeRF first extracts features from the occupancy grid in NeRF. Subsequently, our DReg-NeRF utilizes a transformer architecture with self-attention and cross-attention layers to learn the relations between pairwise NeRF blocks. In contrast to state-of-the-art (SOTA) point cloud registration methods, the decoupled correspondences are supervised by surface fields without any ground truth overlapping labels. We construct a novel view synthesis dataset with 1,700+ 3D objects obtained from Objaverse to train our network. When evaluated on the test set, our proposed method beats the SOTA point cloud registration methods by a large margin with a mean RPE = 9.67° and a mean RTE = 0.038. Our code is available at https://github.com/AIBluefisher/DReg-NeRF.

Real-world data tends to follow a long-tailed distribution, where the class imbalance results in dominance of the head classes during training. In this paper, we propose a frustratingly simple but effective step-wise learning framework to gradually enhance the capability of the model in detecting all categories of long-tailed datasets. Specifically, we build smooth-tail data where the long-tailed distribution of categories decays smoothly to correct the bias towards head classes. We pre-train a model on the whole long-tailed data to preserve discriminability between all categories. We then fine-tune the class-agnostic modules of the pre-trained model on the head class dominant replay data to get a head class expert model with improved decision boundaries from all categories. Finally, we train a unified model on the tail class dominant replay data while transferring knowledge from the head class expert model to ensure accurate detection of all categories. Extensive experiments on long-tailed datasets LVIS v0.5 and LVIS v1.0 demonstrate the superior performance of our method, where we can improve the AP with ResNet-50 backbone from 27.0% to 30.3% AP, and especially for the rare categories from 15.5% to 24.9% AP. Our best model using ResNet-101 backbone can achieve 30.7% AP, which suppresses all existing detectors using the same backbone. Our source code is available at https://github.com/dongnana777/Long-tailed-object-detection.

Cross-domain image retrieval has been extensively studied due to its high practical value. In recently proposed unsupervised cross-domain image retrieval methods, efforts are taken to break the data annotation barrier. However, applicability of the model is still confined to domains seen during training. This limitation motivates us to present the first attempt at domain-generalized unsupervised crossdomain image retrieval (DG-UCDIR) aiming at facilitating image retrieval between any two unseen domains in an unsupervised way. To improve domain generalizability of the model, we thus propose a new two-stage domain augmentation technique for diversified training data generation. DG-UCDIR also shares all the challenges present in the unsupervised cross-domain image retrieval, where domain-agnostic and semantic-aware feature representations are supposed to be learned without external supervision. To accomplish this, we introduce a novel crossdomain contrastive learning strategy by utilizing phase image as a proxy to mitigate the domain gap. Extensive experiments are carried out using PACS and DomainNet dataset, and consistently illustrate the superior performance of our framework compared to existing state-ofthe-art methods. Our source code is available at https://github.com/conghui1002/DG-UCDIR.

Multi-agent cooperative perception is an increasingly popular topic in the field of autonomous driving, where roadside LiDARs play an essential role. However, how to optimize the placement of roadside LiDARs is a crucial but often overlooked problem. This paper proposes an approach to optimize the placement of roadside LiDARs by selecting optimized positions within the scene for better perception performance. To efficiently obtain the best combination of locations, a greedy algorithm based on perceptual gain is proposed, which selects the location that can maximize the perceptual gain sequentially. We define perceptual gain as the increased perceptual capability when a new LiDAR is placed. To obtain the perception capability, we propose a perception predictor that learns to evaluate LiDAR placement using only a single point cloud frame. A dataset named Roadside-Opt is created using the CARLA simulator to facilitate research on the roadside LiDAR placement problem. Extensive experiments are conducted to demonstrate the effectiveness of our proposed method.

Event cameras offer many advantages over standard cameras due to their distinctive principle of operation: low power, low latency, high temporal resolution and high dynamic range. Nonetheless, the success of many downstream visual applications also hinges on an efficient and effective scene representation, where Neural Radiance Field (NeRF) is seen as the leading candidate. Such promise and potential of event cameras and NeRF inspired recent works to investigate on the reconstruction of NeRF from moving event cameras. However, these works are mainly limited in terms of the dependence on dense and low-noise event streams, as well as generalization to arbitrary contrast threshold values and camera speed profiles. In this work, we propose Robust e-NeRF, a novel method to directly and robustly reconstruct NeRFs from moving event cameras under various real-world conditions, especially from sparse and noisy events generated under non-uniform motion. It consists of two key components: a realistic event generation model that accounts for various intrinsic parameters (e.g. timeindependent, asymmetric threshold and refractory period) and non-idealities (e.g. pixel-to-pixel threshold variation), as well as a complementary pair of normalized reconstruction losses that can effectively generalize to arbitrary speed profiles and intrinsic parameter values without such prior knowledge. Experiments on real and novel realistically simulated sequences verify our effectiveness. Our code, synthetic dataset and improved event simulator are public.

Existing fully-supervised point cloud segmentation methods suffer in the dynamic testing environment with emerging new classes. Few-shot point cloud segmentation algorithms address this problem by learning to adapt to new classes at the sacrifice of segmentation accuracy for the base classes, which severely impedes its practicality. This largely motivates us to present the first attempt at a more practical paradigm of generalized few-shot point cloud segmentation, which requires the model to generalize to new categories with only a few support point clouds and simultaneously retain the capability to segment base classes. We propose the geometric words to represent geometric components shared between the base and novel classes, and incorporate them into a novel geometric-aware semantic representation to facilitate better generalization to the new classes without forgetting the old ones. Moreover, we introduce geometric prototypes to guide the segmentation with geometric prior knowledge. Extensive experiments on S3DIS and ScanNet consistently illustrate the superior performance of our method over baseline methods. Our code is available at: https://github.com/Pixie8888/GFS-3DSeg_GWs.

Domain adaptation (DA) aims to transfer knowledge from a fully labeled source to a scarcely labeled or totally unlabeled target under domain shift. Recently, semi-supervised learning-based (SSL) techniques that leverage pseudo labeling have been increasingly used in DA. Despite the competitive performance, these pseudo labeling methods rely heavily on the source domain to generate pseudo labels for the target domain and therefore still suffer considerably from source data bias. Moreover, class distribution bias in the target domain is also often ignored in the pseudo label generation and thus leading to further deterioration of performance. In this paper, we propose GeT that learns a non-bias target embedding distribution with high quality pseudo labels. Specifically, we formulate an online target generative classifier to induce the target distribution into distinctive Gaussian components weighted by their class priors to mitigate source data bias and enhance target class discriminability. We further propose a structure similarity regularization framework to alleviate target class distribution bias and further improve target class discriminability. Experimental results show that our proposed GeT is effective and achieves consistent improvements under various DA settings with and without class distribution bias. Our code is available at: https://lulusindazc.github.io/getproject/.

Despite the impressive results achieved by many existing Structure from Motion (SfM) approaches, there is still a need to improve the robustness, accuracy, and efficiency on large-scale scenes with many outlier matches and sparse view graphs. In this paper, we propose AdaSfM: a coarse-to-fine adaptive SfM approach that is scalable to large-scale and challenging datasets. Our approach first does a coarse global SfM which improves the reliability of the view graph by leveraging measurements from low-cost sensors such as Inertial Measurement Units (IMUs) and wheel encoders. Subsequently, the view graph is divided into sub-scenes that are refined in parallel by a fine local incremental SfM regularised by the result from the coarse global SfM to improve the camera registration accuracy and alleviate scene drifts. Finally, our approach uses a threshold-adaptive strategy to align all local reconstructions to the coordinate frame of global SfM. Extensive experiments on large-scale benchmark datasets show that our approach achieves state-of-the-art accuracy and efficiency. [Project Page]

3D scene segmentation based on neural implicit representation has emerged recently with the advantage of training only on 2D supervision. However, existing approaches still requires expensive per-scene optimization that prohibits generalization to novel scenes during inference. To circumvent this problem, we introduce a \textit{generalizable} 3D segmentation framework based on implicit representation. Specifically, our framework takes in multi-view image features and semantic maps as the inputs instead of only spatial information to avoid overfitting to scene-specific geometric and semantic information. We propose a novel soft voting mechanism to aggregate the 2D semantic information from different views for each 3D point. In addition to the image features, view difference information is also encoded in our framework to predict the voting scores. Intuitively, this allows the semantic information from nearby views to contribute more compared to distant ones. Furthermore, a visibility module is also designed to detect and filter out detrimental information from occluded views. Due to the generalizability of our proposed method, we can synthesize semantic maps or conduct 3D semantic segmentation for novel scenes with solely 2D semantic supervision. Experimental results show that our approach achieves comparable performance with scene-specific approaches. More importantly, our approach can even outperform existing strong supervision-based approaches with only 2D annotations.

Remarkable progress has been made in 3D reconstruction from single-view RGB-D inputs. MCC is the current state-of-the-art method in this field, which achieves unprecedented success by combining vision Transformers with large-scale training. However, we identified two key limitations of MCC: 1) The Transformer decoder is inefficient in handling large number of query points; 2) The 3D representation struggles to recover high-fidelity details. In this paper, we propose a new approach called NU-MCC that addresses these limitations. NU-MCC includes two key innovations: a Neighborhood decoder and a Repulsive Unsigned Distance Function (Repulsive UDF). First, our Neighborhood decoder introduces center points as an efficient proxy of input visual features, allowing each query point to only attend to a small neighborhood. This design not only results in much faster inference speed but also enables the exploitation of finer-scale visual features for improved recovery of 3D textures. Second, our Repulsive UDF is a novel alternative to the occupancy field used in MCC, significantly improving the quality of 3D object reconstruction. Compared to standard UDFs that suffer from holes in results, our proposed Repulsive UDF can achieve more complete surface reconstruction. Experimental results demonstrate that NU-MCC is able to learn a strong 3D representation, significantly advancing the state of the art in single-view 3D reconstruction. Particularly, it outperforms MCC by 9.7% in terms of the F1-score on the CO3D-v2 dataset with more than 5x faster running speed.

Accurately predicting object depth is a key challenge in monocular 3D detection task. The perspective projection principle used by most state-of-the-art approaches demands a complex balance between the ratio-form depth estimation and 2D-3D geometric regularizations, and thus can lead to sub-optimal solutions. In this paper, we propose a novel synergistic scheme that can achieve better trade-off among these competing objectives. Our main proposal is a progressive depth regularization (PDR) architecture that splits the overall training process into three sequential depth estimation steps to gradually remove the unwanted deviations induced by the over-regularization. Specifically, our model first learns the coarse depth with the conventional perspective projection and combines the coarse-to-fine generation to reduce the search space of 2D projection height prediction. We then deactivate individual supervision on 2D projection height prediction and introduces a new auxiliary 3D physical height prediction to relax the 2D and 3D regularizations, respectively. Consequently, our PDR leads to more precise depth estimation by mitigating the inherent ambiguities in the geometric priors of perspective projection through progressive regularization relaxation. Extensive experiments on both KITTI and Rope3D benchmark show that our PDR delivers strong performance gains as compared to the previous methods.

Deep learning-based approaches have shown remarkable performance in the 3D object detection task. However, they suffer from a catastrophic performance drop on the originally trained classes when incrementally learning new classes without revisiting the old data. This "catastrophic forgetting" phenomenon impedes the deployment of 3D object detection approaches in real-world scenarios, where continuous learning systems are needed. In this paper, we study the unexplored yet important class-incremental 3D object detection problem and present the first solution - SDCoT, a novel static-dynamic co-teaching method. Our SDCoT alleviates the catastrophic forgetting of old classes via a static teacher, which provides pseudo annotations for old classes in the new samples and regularizes the current model by extracting previous knowledge with a distillation loss. At the same time, SDCoT consistently learns the underlying knowledge from new data via a dynamic teacher. We conduct extensive experiments on two benchmark datasets and demonstrate the superior performance of our SDCoT over baseline approaches in several incremental learning scenarios. Our code is available at https://github.com/Na-Z/SDCoT.

Despite recent success in incorporating learning into point cloud registration, many works focus on learning feature descriptors and continue to rely on nearest-neighbor feature matching and outlier filtering through RANSAC to obtain the final set of correspondences for pose estimation. In this work, we conjecture that attention mechanisms can replace the role of explicit feature matching and RANSAC, and thus propose an end-to-end framework to directly predict the final set of correspondences. We use a network architecture consisting primarily of transformer layers containing self and cross attentions, and train it to predict the probability each point lies in the overlapping region and its corresponding position in the other point cloud. The required rigid transformation can then be estimated directly from the predicted correspondences without further post-processing. Despite its simplicity, our approach achieves state-of-the-art performance on 3DMatch and ModelNet benchmarks. Our source code can be found at https://github.com/yewzijian/RegTR.

We introduce a new setting of Novel Class Discovery in Semantic Segmentation (NCDSS), which aims at segmenting unlabeled images containing new classes given prior knowledge from a labeled set of disjoint classes. In contrast to existing approaches that look at novel class dis-covery in image classification, we focus on the more chal-lenging semantic segmentation. In NCDSS, we need to dis-tinguish the objects and background, and to handle the existence of multiple classes within an image, which in-creases the difficulty in using the unlabeled data. To tackle this new setting, we leverage the labeled base data and a saliency model to coarsely cluster novel classes for model training in our basic framework. Additionally, we propose the Entropy-based Uncertainty Modeling and Self-training (EUMS) framework to overcome noisy pseudo-labels, fur-ther improving the model performance on the novel classes. Our EUMS utilizes an entropy ranking technique and a dy-namic reassignment to distill clean labels, thereby making full use of the noisy data via self-supervised learning. We build the NCDSS benchmark on the PASCAL-5idataset and COCO-20i dataset. Extensive experiments demonstrate the feasibility of the basic framework (achieving an average mIoU of 49.81% on PASCAL-5i) and the effectiveness of EUMS framework (outperforming the basic framework by 9.28% mIoU on PASCAL-5i).

In this paper, we consider the problem of domain generalization in semantic segmentation, which aims to learn a robust model using only labeled synthetic (source) data. The model is expected to perform well on unseen real (target) domains. Our study finds that the image style variation can largely influence the model's performance and the style features can be well represented by the channel-wise mean and standard deviation of images. Inspired by this, we propose a novel adversarial style augmentation (AdvStyle) approach, which can dynamically generate hard stylized images during training and thus can effectively prevent the model from overfitting on the source domain. Specifically, AdvStyle regards the style feature as a learnable parameter and updates it by adversarial training. The learned adversarial style feature is used to construct an adversarial image for robust model training. AdvStyle is easy to implement and can be readily applied to different models. Experiments on two synthetic-to-real semantic segmentation benchmarks demonstrate that AdvStyle can significantly improve the model performance on unseen real domains and show that we can achieve the state of the art. Moreover, AdvStyle can be employed to domain generalized image classification and produces a clear improvement on the considered datasets.

Point clouds are often sparse and incomplete, which imposes difficulties for real-world applications. Existing shape completion methods tend to generate rough shapes without fine-grained details. Considering this, we introduce a two-branch network for shape completion. The first branch is a cascaded shape completion sub-network to synthesize complete objects, where we propose to use the partial input together with the coarse output to preserve the object details during the dense point reconstruction. The second branch is an auto-encoder to reconstruct the original partial input. The two branches share a same feature extractor to learn an accurate global feature for shape completion. Furthermore, we propose two strategies to enable the training of our network when ground truth data are not available. This is to mitigate the dependence of existing approaches on large amounts of ground truth training data that are often difficult to obtain in real-world applications. Additionally, our proposed strategies are also able to improve the reconstruction quality for fully supervised learning. We verify our approach in self-supervised, semi-supervised and fully supervised settings with superior performances. Quantitative and qualitative results on different datasets demonstrate that our method achieves more realistic outputs than state-of-the-art approaches on the point cloud completion task.

In this work, we introduce a new concept, named source-free open compound domain adaptation (SF-OCDA), and study it in semantic segmentation. SF-OCDA is more challenging than the traditional domain adaptation but it is more practical. It jointly considers (1) the issues of data privacy and data storage and (2) the scenario of multiple target domains and unseen open domains. In SF-OCDA, only the source pre-trained model and the target data are available to learn the target model. The model is evaluated on the samples from the target and unseen open domains. To solve this problem, we present an effective framework by separating the training process into two stages: (1) pre-training a generalized source model and (2) adapting a target model with self-supervised learning. In our framework, we propose the Cross-Patch Style Swap (CPSS) to diversify samples with various patch styles in the feature-level, which can benefit the training of both stages. First, CPSS can significantly improve the generalization ability of the source model, providing more accurate pseudo-labels for the latter stage. Second, CPSS can reduce the influence of noisy pseudo-labels and also avoid the model overfitting to the target domain during self-supervised learning, consistently boosting the performance on the target and open domains. Experiments demonstrate that our method produces state-of-the-art results on the C-Driving dataset. Furthermore, our model also achieves the leading performance on CityScapes for domain generalization.

Transformation Synchronization is the problem of recovering absolute transformations from a given set of pairwise relative motions. Despite its usefulness, the problem remains challenging due to the influences from noisy and outlier relative motions, and the difficulty to model analytically and suppress them with high fidelity. In this work, we avoid handcrafting robust loss functions, and propose to use graph neural networks (GNNs) to learn transformation synchronization. Unlike previous works which use complicated multi-stage pipelines, we use an iterative approach where each step consists of a single weight-shared message passing layer that refines the absolute poses from the previous iteration by predicting an incremental update in the tangent space. To reduce the influence of outliers, the messages are weighted before aggregation. Our iterative approach alleviates the need for an explicit initialization step and performs well with identity initial poses. Although our approach is simple, we show that it performs favorably against existing handcrafted and learned synchronization methods through experiments on both SO(3) and SE(3) synchronization.

Animal pose estimation is an important field that has received increasing attention in the recent years. The main challenge for this task is the lack of labeled data. Existing works circumvent this problem with pseudo labels generated from data of other easily accessible domains such as synthetic data. However, these pseudo labels are noisy even with consistency check or confidence-based filtering due to the domain shift in the data. To solve this problem, we design a multi-scale domain adaptation module (MDAM) to reduce the domain gap between the synthetic and real data. We further introduce an online coarse-to-fine pseudo label updating strategy. Specifically, we propose a self-distillation module in an inner coarse-update loop and a mean-teacher in an outer fine-update loop to generate new pseudo labels that gradually replace the old ones. Consequently, our model is able to learn from the old pseudo labels at the early stage, and gradually switch to the new pseudo labels to prevent overfitting in the later stage. We evaluate our approach on the TigDog and VisDA 2019 datasets, where we outperform existing approaches by a large margin. We also demonstrate the generalization ability of our model by testing extensively on both unseen domains and unseen animal categories. Our code is available at the project website1.

This paper presents DeepI2P: a novel approach for cross-modality registration between an image and a point cloud. Given an image (e.g. from a rgb-camera) and a general point cloud (e.g. from a 3D Lidar scanner) captured at different locations in the same scene, our method estimates the relative rigid transformation between the co-ordinate frames of the camera and Lidar. Learning common feature descriptors to establish correspondences for the registration is inherently challenging due to the lack of appearance and geometric correlations across the two modalities. We circumvent the difficulty by converting the registration problem into a classification and inverse camera projection optimization problem. A classification neural network is designed to label whether the projection of each point in the point cloud is within or beyond the camera frustum. These labeled points are subsequently passed into a novel inverse camera projection solver to estimate the relative pose. Extensive experimental results on Oxford Robotcar and KITTI datasets demonstrate the feasibility of our approach. Our source code is available at https://github.com/lijx10/DeepI2P.

Existing approaches for multi-view multi-person 3D pose estimation explicitly establish cross-view correspondences to group 2D pose detections from multiple camera views and solve for the 3D pose estimation for each person. Establishing cross-view correspondences is challenging in multi-person scenes, and incorrect correspondences will lead to sub-optimal performance for the multi-stage pipeline. In this work, we present our multi-view 3D pose estimation approach based on plane sweep stereo to jointly address the cross-view fusion and 3D pose reconstruction in a single shot. Specifically, we propose to perform depth regression for each joint of each 2D pose in a target camera view. Cross-view consistency constraints are implicitly enforced by multiple reference camera views via the plane sweep algorithm to facilitate accurate depth regression. We adopt a coarse-to-fine scheme to first regress the person-level depth followed by a per-person joint-level relative depth estimation. 3D poses are obtained from a simple back-projection given the estimated depths. We evaluate our approach on benchmark datasets where it outperforms previous state-of-the-arts while being remarkably efficient. Our code is available at the project website.1

Many existing approaches for 3D point cloud semantic segmentation are fully supervised. These fully supervised approaches heavily rely on large amounts of labeled training data that are difficult to obtain and cannot segment new classes after training. To mitigate these limitations, we propose a novel attention-aware multi-prototype transductive few-shot point cloud semantic segmentation method to segment new classes given a few labeled examples. Specifically, each class is represented by multiple prototypes to model the complex data distribution of labeled points. Subsequently, we employ a transductive label propagation method to exploit the affinities between labeled multi-prototypes and unlabeled points, and among the unlabeled points. Furthermore, we design an attention-aware multi-level feature learning network to learn the discriminative features that capture the geometric dependencies and semantic correlations between points. Our proposed method shows significant and consistent improvements compared to baselines in different few-shot point cloud semantic segmentation settings (i.e. 2/3-way 1/5-shot) on two benchmark datasets. Our code is available at https://github.com/Na-Z/attMPTI.

In this paper, we propose MINE to perform novel view synthesis and depth estimation via dense 3D reconstruction from a single image. Our approach is a continuous depth generalization of the Multiplane Images (MPI) by introducing the NEural radiance fields (NeRF). Given a single image as input, MINE predicts a 4-channel image (RGB and volume density) at arbitrary depth values to jointly reconstruct the camera frustum and fill in occluded contents. The reconstructed and inpainted frustum can then be easily rendered into novel RGB or depth views using differentiable rendering. Extensive experiments on RealEstate10K, KITTI and Flowers Light Fields show that our MINE outperforms state-of-the-art by a large margin in novel view synthesis. We also achieve competitive results in depth estimation on iBims-1 and NYU-v2 without annotated depth supervision. Our source code is available at https://github.com/vincentfung13/MINE.

Deep learning technique has yielded significant improvements in point cloud completion with the aim of completing missing object shapes from partial inputs. However, most existing methods fail to recover realistic structures due to over-smoothing of fine-grained details. In this paper, we develop a voxel-based network for point cloud completion by leveraging edge generation (VE-PCN). We first embed point clouds into regular voxel grids, and then generate complete objects with the help of the hallucinated shape edges. This decoupled architecture together with a multi-scale grid feature learning is able to generate more realistic on-surface details. We evaluate our model on the publicly available completion datasets and show that it outperforms existing state-of-the-art approaches quantitatively and qualitatively. Our source code is available at https://github.com/xiaogangw/VE-PCN.

Steerable CNN imposes the prior knowledge of transformation invariance or equivariance in the network architecture to enhance the the network robustness on geometry transformation of data and reduce overfitting. It has been an intuitive and widely used technique to construct a steerable filter by augmenting a filter with its transformed copies in the past decades, which is named as filter transform in this paper. Recently, the problem of steerable CNN has been studied from aspect of group representation theory, which reveals the function space structure of a steerable kernel function. However, it is not yet clear on how this theory is related to the filter transform technique. In this paper, we show that kernel constructed by filter transform can also be interpreted in the group representation theory. This interpretation help complete the puzzle of steerable CNN theory and provides a novel and simple approach to implement steerable convolution operators. Experiments are executed on multiple datasets to verify the feasibility of the proposed approach.

Bottom-up approaches for image-based multi-person pose estimation consist of two stages: (1) keypoint detection and (2) grouping of the detected keypoints to form person instances. Current grouping approaches rely on learned embedding from only visual features that completely ignore the spatial configuration of human poses. In this work, we formulate the grouping task as a graph partitioning problem, where we learn the affinity matrix with a Graph Neural Network (GNN). More specifically, we design a Geometry-aware Association GNN that utilizes spatial information of the keypoints and learns local affinity from the global context. The learned geometry-based affinity is further fused with appearance-based affinity to achieve robust keypoint association. Spectral clustering is used to partition the graph for the formation of the pose instances. Experimental results on two benchmark datasets show that our proposed method outperforms existing appearance-only grouping frameworks, which shows the effectiveness of utilizing spatial context for robust grouping. Source code is available at: https://github.com/jiahaoLjh/PoseGrouping.

We propose a method for detecting structural changes in a city using images captured from vehicular mounted cameras over traversals at two different times. We first generate 3D point clouds for each traversal from the images and approximate GNSS/INS readings using Structure-from-Motion (SfM). A direct comparison of the two point clouds for change detection is not ideal due to inaccurate geo-location information and possible drifts in the SfM. To circumvent this problem, we propose a deep learning-based non-rigid registration on the point clouds which allows us to compare the point clouds for structural change detection in the scene. Furthermore, we introduce a dual thresholding check and post-processing step to enhance the robustness of our method. We collect two datasets for the evaluation of our approach. Experiments show that our method is able to detect scene changes effectively, even in the presence of viewpoint and illumination differences.

The environment of most real-world scenarios such as malls and supermarkets changes at all times. A pre-built map that does not account for these changes becomes out-of-date easily. Therefore, it is necessary to have an up-to-date model of the environment to facilitate long-term operation of a robot. To this end, this paper presents a general lifelong simultaneous localization and mapping (SLAM) framework. Our framework uses a multiple session map representation, and exploits an efficient map updating strategy that includes map building, pose graph refinement and sparsification. To mitigate the unbounded increase of memory usage, we propose a map-trimming method based on the Chow-Liu maximum-mutual-information spanning tree. The proposed SLAM framework has been comprehensively validated by over a month of robot deployment in real supermarket environment. Furthermore, we release the dataset collected from the indoor and outdoor changing environment with the hope to accelerate lifelong SLAM research in the community. Our dataset is available at https://github.com/sanduan168/lifelong-SLAM-dataset.

Deep networks have shown remarkable results in the task of object detection. However, their performance suffers critical drops when they are subsequently trained on novel classes without any sample from the base classes originally used to train the model. This phenomenon is known as catastrophic forgetting. Recently, several incremental learning methods are proposed to mitigate catastrophic forgetting for object detection. Despite the effectiveness, these methods require co-occurrence of the unlabeled base classes in the training data of the novel classes. This requirement is impractical in many real-world settings since the base classes do not necessarily co-occur with the novel classes. In view of this limitation, we consider a more practical setting of complete absence of co-occurrence of the base and novel classes for the object detection task. We propose the use of unlabeled in-the-wild data to bridge the non co-occurrence caused by the missing base classes during the training of additional novel classes. To this end, we introduce a blind sampling strategy based on the responses of the base-class model and pre-trained novel-class model to select a smaller relevant dataset from the large in-the-wild dataset for incremental learning. We then design a dual-teacher distillation framework to transfer the knowledge distilled from the base- and novel-class teacher models to the student model using the sampled in-the-wild data. Experimental results on the PASCAL VOC and MS COCO datasets show that our proposed method significantly outperforms other state-of-the-art class-incremental object detection methods when there is no co-occurrence between the base and novel classes during training.

Most existing animal pose and shape estimation approaches reconstruct animal meshes with a parametric SMAL model. This is because the low-dimensional pose and shape parameters of the SMAL model makes it easier for deep networks to learn the high-dimensional animal meshes. However, the SMAL model is learned from scans of toy animals with limited pose and shape variations, and thus may not be able to represent highly varying real animals well. This may result in poor fittings of the estimated meshes to the 2D evidences, e.g. 2D keypoints or silhouettes. To mitigate this problem, we propose a coarse-to-fine approach to reconstruct 3D animal mesh from a single image. The coarse estimation stage first estimates the pose, shape and translation parameters of the SMAL model. The estimated meshes are then used as a starting point by a graph convolutional network (GCN) to predict a per-vertex deformation in the refinement stage. This combination of SMAL-based and vertex-based representations benefits from both parametric and non-parametric representations. We design our mesh refinement GCN (MRGCN) as an encoder-decoder structure with hierarchical feature representations to overcome the limited receptive field of traditional GCNs. Moreover, we observe that the global image feature used by existing animal mesh reconstruction works is unable to capture detailed shape information for mesh refinement. We thus introduce a local feature extractor to retrieve a vertex-level feature and use it together with the global feature as the input of the MRGCN. We test our approach on the StanfordExtra dataset and achieve state-of-the-art results. Furthermore, we test the generalization capacity of our approach on the Animal Pose and BADJA datasets. Our code is available at the project website.

Point clouds are often sparse and incomplete. Existing shape completion methods are incapable of generating details of objects or learning the complex point distributions. To this end, we propose a cascaded refinement network together with a coarse-to-fine strategy to synthesize the detailed object shapes. Considering the local details of partial input with the global shape information together, we can preserve the existing details in the incomplete point set and generate the missing parts with high fidelity. We also design a patch discriminator that guarantees every local area has the same pattern with the ground truth to learn the complicated point distribution. Quantitative and qualitative experiments on different datasets show that our method achieves superior results compared to existing state-of-the-art approaches on the 3D point cloud completion task. Our source code is available at https://github.com/xiaogangw/cascaded-point-completion.git.

Point cloud analysis has received much attention recently; and segmentation is one of the most important tasks. The success of existing approaches is attributed to deep network design and large amount of labelled training data, where the latter is assumed to be always available. However, obtaining 3d point cloud segmentation labels is often very costly in practice. In this work, we propose a weakly supervised point cloud segmentation approach which requires only a tiny fraction of points to be labelled in the training stage. This is made possible by learning gradient approximation and exploitation of additional spatial and color smoothness constraints. Experiments are done on three public datasets with different degrees of weak supervision. In particular, our proposed method can produce results that are close to and sometimes even better than its fully supervised counterpart with 10× fewer labels. Our code is available at the project website1.

Iterative Closest Point (ICP) solves the rigid point cloud registration problem iteratively in two steps: (1) make hard assignments of spatially closest point correspondences, and then (2) find the least-squares rigid transformation. The hard assignments of closest point correspondences based on spatial distances are sensitive to the initial rigid transformation and noisy/outlier points, which often cause ICP to converge to wrong local minima. In this paper, we propose the RPM-Net - a less sensitive to initialization and more robust deep learning-based approach for rigid point cloud registration. To this end, our network uses the differentiable Sinkhorn layer and annealing to get soft assignments of point correspondences from hybrid features learned from both spatial coordinates and local geometry. To further improve registration performance, we introduce a secondary network to predict optimal annealing parameters. Unlike some existing methods, our RPM-Net handles missing correspondences and point clouds with partial visibility. Experimental results show that our RPM-Net achieves state-of-the-art performance compared to existing non-deep learning and recent deep learning methods. Our source code is available at the project website (https://github.com/yewzijian/RPMNet).

The performance of existing point cloud-based 3D object detection methods heavily relies on large-scale high-quality 3D annotations. However, such annotations are often tedious and expensive to collect. Semi-supervised learning is a good alternative to mitigate the data annotation issue, but has remained largely unexplored in 3D object detection. Inspired by the recent success of self-ensembling technique in semi-supervised image classification task, we propose SESS, a self-ensembling semi-supervised 3D object detection framework. Specifically, we design a thorough perturbation scheme to enhance generalization of the network on unlabeled and new unseen data. Furthermore, we propose three consistency losses to enforce the consistency between two sets of predicted 3D object proposals, to facilitate the learning of structure and semantic invariances of objects. Extensive experiments conducted on SUN RGB-D and ScanNet datasets demonstrate the effectiveness of SESS in both inductive and transductive semi-supervised 3D object detection. Our SESS achieves competitive performance compared to the state-of-the-art fully-supervised method by using only 50% labeled data. Our code is available at https://github.com/Na-Z/sess.

In this paper, we present the PS2-Net - a locally and globally aware deep learning framework for semantic segmentation on 3D scene-level point clouds. In order to deeply incorporate local structures and global context to support 3D scene segmentation, our network is built on four repeatedly stacked encoders, where each encoder has two basic components: EdgeConv that captures local structures and NetVLAD that models global context. Different from existing state-of-the-art methods for point-based scene semantic segmentation that either violate or do not achieve permutation invariance, our PS2-Net is designed to be permutation invariant which is an essential property of any deep network used to process unordered point clouds. We further provide theoretical proof to guarantee the permutation invariance property of our network. We perform extensive experiments on two large-scale 3D indoor scene datasets and demonstrate that our PS2-Net is able to achieve state-of-the-art performances as compared to existing approaches.

Motivated by the success of encoding multi-scale contextual information for image analysis, we propose our PointAtrousGraph (PAG) - a deep permutation-invariant hierarchical encoder-decoder for efficiently exploiting multi-scale edge features in point clouds. Our PAG is constructed by several novel modules, such as Point Atrous Convolution (PAC), Edgepreserved Pooling (EP) and Edge-preserved Unpooling (EU). Similar with atrous convolution, our PAC can effectively enlarge receptive fields of filters and thus densely learn multi-scale point features. Following the idea of non-overlapping maxpooling operations, we propose our EP to preserve critical edge features during subsampling. Correspondingly, our EU modules gradually recover spatial information for edge features. In addition, we introduce chained skip subsampling/upsampling modules that directly propagate edge features to the final stage. Particularly, our proposed auxiliary loss functions can further improve our performance. Experimental results show that our PAG outperform previous state-of-the-art methods on various 3D semantic perception applications.

Accurate 6D object pose estimation is fundamental to robotic manipulation and grasping. Previous methods follow a local optimization approach which minimizes the distance between closest point pairs to handle the rotation ambiguity of symmetric objects. In this work, we propose a novel discrete- continuous formulation for rotation regression to resolve this local-optimum problem. We uniformly sample rotation anchors in SO(3), and predict a constrained deviation from each anchor to the target, as well as uncertainty scores for selecting the best prediction. Additionally, the object location is detected by aggregating point-wise vectors pointing to the 3D center. Experiments on two benchmarks: LINEMOD and YCB-Video, show that the proposed method outperforms state-of-the-art approaches. Our code is available at https://github.com/mentian/object-posenet.

In view of the difficulty in reconstructing object details in point cloud completion, we propose a shape prior learning method for object completion. The shape priors include geometric information in both complete and the partial point clouds. We design a feature alignment strategy to learn the shape prior from complete points, and a coarse to fine strategy to incorporate partial prior in the fine stage. To learn the complete objects prior, we first train a point cloud auto-encoder to extract the latent embeddings from complete points. Then we learn a mapping to transfer the point features from partial points to that of the complete points by optimizing feature alignment losses. The feature alignment losses consist of a L2 distance and an adversarial loss obtained by Maximum Mean Discrepancy Generative Adversarial Network (MMD-GAN). The L2 distance optimizes the partial features towards the complete ones in the feature space, and MMD-GAN decreases the statistical distance of two point features in a Reproducing Kernel Hilbert Space. We achieve state-of-the-art performances on the point cloud completion task. Our code is available at https://github.com/xiaogangw/point-cloud-completion-shape-prior.

Outlier feature matches and loop-closures that survived front-end data association can lead to catastrophic failures in the back-end optimization of large-scale point cloud based 3D reconstruction. To alleviate this problem, we propose a probabilistic approach for robust back-end optimization in the presence of outliers. More specifically, we model the problem as a Bayesian network and solve it using the Expectation-Maximization algorithm. Our approach leverages on a long-tail Cauchy distribution to suppress outlier feature matches in the odometry constraints, and a Cauchy-Uniform mixture model with a set of binary latent variables to simultaneously suppress outlier loop-closure constraints and outlier feature matches in the inlier loop-closure constraints. Furthermore, we show that by using a Gaussian-Uniform mixture model, our approach degenerates to the formulation of a state-of-the-art approach for robust indoor reconstruction. Experimental results demonstrate that our approach has comparable performance with the state-of-the-art on a benchmark indoor dataset, and outperforms it on a large-scale outdoor dataset. Our source code can be found on the project website.

3D human pose estimation from a monocular image or 2D joints is an ill-posed problem because of depth ambiguity and occluded joints. We argue that 3D human pose estimation from a monocular input is an inverse problem where multiple feasible solutions can exist. In this paper, we propose a novel approach to generate multiple feasible hypotheses of the 3D pose from 2D joints. In contrast to existing deep learning approaches which minimize a mean square error based on an unimodal Gaussian distribution, our method is able to generate multiple feasible hypotheses of 3D pose based on a multimodal mixture density networks. Our experiments show that the 3D poses estimated by our approach from an input of 2D joints are consistent in 2D reprojections, which supports our argument that multiple solutions exist for the 2D-to-3D inverse problem. Furthermore, we show state-of-the-art performance on the Human3.6M dataset in both best hypothesis and multi-view settings, and we demonstrate the generalization capacity of our model by testing on the MPII and MPI-INF-3DHP datasets. Our code is available at the project website.

In this paper, we propose the USIP detector: an Unsupervised Stable Interest Point detector that can detect highly repeatable and accurately localized keypoints from 3D point clouds under arbitrary transformations without the need for any ground truth training data. Our USIP detector consists of a feature proposal network that learns stable keypoints from input 3D point clouds and their respective transformed pairs from randomly generated transformations. We provide degeneracy analysis and suggest solutions to prevent it. We encourage high repeatability and accurate localization of the keypoints with a probabilistic chamfer loss that minimizes the distances between the detected keypoints from the training point cloud pairs. Extensive experimental results of repeatability tests on several simulated and real-world 3D point cloud datasets from Lidar, RGB-D and CAD models show that our USIP detector significantly outperforms existing hand-crafted and deep learning-based 3D keypoint detectors. Our code is available at the project website. https://github.com/lijx10/USIP.

We investigate the direction of training a 3D object detector for new object classes from only 2D bounding box labels of these new classes, while simultaneously transferring information from 3D bounding box labels of the existing classes. To this end, we propose a transferable semi-supervised 3D object detection model that learns a 3D object detector network from training data with two disjoint sets of object classes - a set of strong classes with both 2D and 3D box labels, and another set of weak classes with only 2D box labels. In particular, we suggest a relaxed reprojection loss, box prior loss and a Box-to-Point Cloud Fit network that allow us to effectively transfer useful 3D information from the strong classes to the weak classes during training, and consequently, enable the network to detect 3D objects in the weak classes during inference. Experimental results show that our proposed algorithm outperforms baseline approaches and achieves promising results compared to fully-supervised approaches on the SUN-RGBD and KITTI datasets. Furthermore, we show that our Box-to-Point Cloud Fit network improves performances of the fully-supervised approaches on both datasets.

Large-scale point cloud generated from 3D sensors is more accurate than its image-based counterpart. However, it is seldom used in visual pose estimation due to the difficulty in obtaining 2D-3D image to point cloud correspondences. In this paper, we propose the 2D3D-MatchNet - an end-to-end deep network architecture to jointly learn the descriptors for 2D and 3D keypoint from image and point cloud, respectively. As a result, we are able to directly match and establish 2D-3D correspondences from the query image and 3D point cloud reference map for visual pose estimation. We create our Oxford 2D-3D Patches dataset from the Oxford Robotcar dataset with the ground truth camera poses and 2D-3D image to point cloud correspondences for training and testing the deep network. Experimental results verify the feasibility of our approach.

Project AutoVision aims to develop localization and 3D scene perception capabilities for a self-driving vehicle. Such capabilities will enable autonomous navigation in urban and rural environments, in day and night, and with cameras as the only exteroceptive sensors. The sensor suite employs many cameras for both 360-degree coverage and accurate multi-view stereo; the use of low-cost cameras keeps the cost of this sensor suite to a minimum. In addition, the project seeks to extend the operating envelope to include GNSS-less conditions which are typical for environments with tall buildings, foliage, and tunnels. Emphasis is placed on leveraging multi-view geometry and deep learning to enable the vehicle to localize and perceive in 3D space. This paper presents an overview of the project, and describes the sensor suite and current progress in the areas of calibration, localization, and perception.

Despite the recent active research on processing point clouds with deep networks, few attention has been on the sensitivity of the networks to rotations. In this paper, we propose a deep learning architecture that achieves discrete SO(2)/SO(3) rotation equivariance for point cloud recognition. Specifically, the rotation of an input point cloud with elements of a rotation group is similar to shuffling the feature vectors generated by our approach. The equivariance is easily reduced to invariance by eliminating the permutation with operations such as maximum or average. Our method can be directly applied to any existing point cloud based networks, resulting in significant improvements in their performance for rotated inputs. We show state-of-the-art results in the classification tasks with various datasets under both SO(2) and SO(3) rotations. In addition, we further analyze the necessary conditions of applying our approach to PointNet [1] based networks.

Self-calibration of camera intrinsics and radial distortion has a long history of research in the computer vision community. However, it remains rare to see real applications of such techniques to modern Simultaneous Localization And Mapping (SLAM) systems, especially in driving scenarios. In this paper, we revisit the geometric approach to this problem, and provide a theoretical proof that explicitly shows the ambiguity between radial distortion and scene depth when two-view geometry is used to self-calibrate the radial distortion. In view of such geometric degeneracy, we propose a learning approach that trains a convolutional neural network (CNN) on a large amount of synthetic data. We demonstrate the utility of our proposed method by applying it as a checkerboard-free calibration tool for SLAM, achieving comparable or superior performance to previous learning and hand-crafted methods.

Image-based localization plays an important role in today's autonomous driving technologies. However, in large scale outdoor environments, challenging conditions, e.g., lighting changes or different weather, heavily affect image appearance and quality. As a key component of feature-based visual localization, image feature detection and matching deteriorate severely and cause worse localization performance. In this paper, we propose a novel method for robust image feature matching under drastically changing outdoor environments. In contrast to existing approaches which try to learn robust feature descriptors, we train a deep network that outputs the low-rank representations of the images where the undesired variations on the images are removed, and perform feature extraction and matching on the learned low-rank space. We demonstrate that our learned low-rank images largely improve the performance of image feature matching under varying conditions over a long period of time.

The problem of localization on a geo-referenced aerial/satellite map given a query ground view image remains challenging due to the drastic change in viewpoint that causes traditional image descriptors based matching to fail. We leverage on the recent success of deep learning to propose the CVM-Net for the cross-view image-based ground-to-aerial geo-localization task. Specifically, our network is based on the Siamese architecture to do metric learning for the matching task. We first use the fully convolutional layers to extract local image features, which are then encoded into global image descriptors using the powerful NetVLAD. As part of the training procedure, we also introduce a simple yet effective weighted soft margin ranking loss function that not only speeds up the training convergence but also improves the final matching accuracy. Experimental results show that our proposed network significantly outperforms the state-of-the-art approaches on two existing benchmarking datasets. Our code and models are publicly available on the project website.

Drone-mounted flying cameras will revolutionize photo-taking. The user, instead of holding a camera in hand and manually searching for a viewpoint, will interact directly with image contents in the viewfinder through simple gestures, and the flying camera will achieve the desired viewpoint through the autonomous flying capability of the drone. This work studies a common situation in photo-taking, i.e., the underlying viewpoint search problem for composing a photo with two objects of interest. We model it as a Perspective-2-Point (P2P) problem, which is under-constrained to determine the six degrees-of-freedom camera pose uniquely. By incorporating the user's composition requirements and minimizing the camera's flying distance, we form a constrained nonlinear optimization problem and solve it in closed form. Experiments on synthetic data sets and on a flying camera system indicate promising results.

This paper presents SO-Net, a permutation invariant architecture for deep learning with orderless point clouds. The SO-Net models the spatial distribution of point cloud by building a Self-Organizing Map (SOM). Based on the SOM, SO-Net performs hierarchical feature extraction on individual points and SOM nodes, and ultimately represents the input point cloud by a single feature vector. The receptive field of the network can be systematically adjusted by conducting point-to-node k nearest neighbor search. In recognition tasks such as point cloud reconstruction, classification, object part segmentation and shape retrieval, our proposed network demonstrates performance that is similar with or better than state-of-the-art approaches. In addition, the training speed is significantly faster than existing point cloud recognition networks because of the parallelizability and simplicity of the proposed architecture. Our code is available at the project website1.

Human motion modeling is a classic problem in computer vision and graphics. Challenges in modeling human motion include high dimensional prediction as well as extremely complicated dynamics.We present a novel approach to human motion modeling based on convolutional neural networks (CNN). The hierarchical structure of CNN makes it capable of capturing both spatial and temporal correlations effectively. In our proposed approach, a convolutional long-term encoder is used to encode the whole given motion sequence into a long-term hidden variable, which is used with a decoder to predict the remainder of the sequence. The decoder itself also has an encoder-decoder structure, in which the short-term encoder encodes a shorter sequence to a short-term hidden variable, and the spatial decoder maps the long and short-term hidden variable to motion predictions. By using such a model, we are able to capture both invariant and dynamic information of human motion, which results in more accurate predictions. Experiments show that our algorithm outperforms the state-of-the-art methods on the Human3.6M and CMU Motion Capture datasets. Our code is available at the project website.

Unlike its image based counterpart, point cloud based retrieval for place recognition has remained as an unexplored and unsolved problem. This is largely due to the difficulty in extracting local feature descriptors from a point cloud that can subsequently be encoded into a global descriptor for the retrieval task. In this paper, we propose the PointNetVLAD where we leverage on the recent success of deep networks to solve point cloud based retrieval for place recognition. Specifically, our PointNetVLAD is a combination/modification of the existing PointNet and NetVLAD, which allows end-to-end training and inference to extract the global descriptor from a given 3D point cloud. Furthermore, we propose the "lazy triplet and quadruplet" loss functions that can achieve more discriminative and generalizable global descriptors to tackle the retrieval task. We create benchmark datasets for point cloud based retrieval for place recognition, and the experimental results on these datasets show the feasibility of our PointNetVLAD. Our code and datasets are publicly available on the project website.

Many existing translation averaging algorithms are either sensitive to disparate camera baselines and have to rely on extensive preprocessing to improve the observed Epipolar Geometry graph, or if they are robust against disparate camera baselines, require complicated optimization to minimize the highly nonlinear angular error objective. In this paper, we carefully design a simple yet effective bilinear objective function, introducing a variable to perform the requisite normalization. The objective function enjoys the baseline-insensitive property of the angular error and yet is amenable to simple and efficient optimization by block coordinate descent, with good empirical performance. A rotation-assisted Iterative Reweighted Least Squares scheme is further put forth to help deal with outliers. We also contribute towards a better understanding of the behavior of two recent convex algorithms, LUD [20] and Shapefit/kick [9], clarifying the underlying subtle difference that leads to the performance gap. Finally, we demonstrate that our algorithm achieves overall superior accuracies in benchmark dataset compared to state-of-the-art methods, and is also several times faster.

While 3D LiDAR has become a common practice for more and more autonomous driving systems, precise 3D mapping and robust localization is of great importance. However, current 3D map is always noisy and unreliable due to the existence of moving objects, leading to worse localization. In this paper, we propose a general vehicle-free point cloud mapping framework for better on-vehicle localization. For each laser scan, vehicle points are detected, tracked and then removed. Simultaneously, 3D map is reconstructed by registering each vehicle-free laser scan to global coordinate based on GPS/INS data. Instead of direct 3D object detection from point cloud, we first detect vehicles from RGB images using the proposed YVDN. In case of false or missing detection, which may result in the existence of vehicles in the map, we propose the K-Frames forward-backward object tracking algorithm to link detection from neighborhood images. Laser scan points falling into the detected bounding boxes are then removed. We conduct our experiments on the Oxford RobotCar Dataset and show the qualitative results to validate the feasibility of our vehicle-free 3D mapping system. Besides, our vehicle-free mapping system can be generalized to any autonomous driving system equipped with LiDAR, camera and/or GPS.

This paper solves the chicken-and-egg problem of line association and vanishing point estimation given a set of lines from an uncalibrated camera image that are corrupted with noise and outliers without making the Manhattan/Atlanta world assumptions. We first build a preference matrix where each column represents the model of a candidate vanishing point and each row indicates the probability of consensus of an image line to the models. We then cast the problem as Binary Quadratic Programming (BQP), where we maximize the selection of dissimilar vanishing point models with large total consensus probability from the preference matrix. We bound the maximization by constraining the maximum number of models that can be selected. Furthermore, we propose a simple yet effective method based on similarity checks between the selected and other models in the preference matrix to decide on the maximum number of models. Experimental results on the open-source benchmark York Urban dataset show the feasibility of our proposed algorithm.

In this paper, we develop a modified differential Structure from Motion (SfM) algorithm that can estimate relative pose from two consecutive frames despite of Rolling Shutter (RS) artifacts. In particular, we show that under constant velocity assumption, the errors induced by the rolling shutter effect can be easily rectified by a linear scaling operation on each optical flow. We further propose a 9-point algorithm to recover the relative pose of a rolling shutter camera that undergoes constant acceleration motion. We demonstrate that the dense depth maps recovered from the relative pose of the RS camera can be used in a RS-aware warping for image rectification to recover high-quality Global Shutter (GS) images. Experiments on both synthetic and real RS images show that our RS-aware differential SfM algorithm produces more accurate results on relative pose estimation and 3D reconstruction from images distorted by RS effect compared to standard SfM algorithms that assume a GS camera model. We also demonstrate that our RS-aware warping for image rectification method outperforms state-of-the-art commercial software products, i.e. Adobe After Effects and Apple Imovie, at removing RS artifacts.

Although 2D LiDAR based Simultaneous Localization and Mapping (SLAM) is a relatively mature topic nowadays, the loop closure problem remains challenging due to the lack of distinctive features in 2D LiDAR range scans. Existing research can be roughly divided into correlation based approaches e.g. scan-to-submap matching and feature based methods e.g. bag-of-words (BoW). In this paper, we solve loop closure detection and relative pose transformation using 2D LiDAR within an end-to-end Deep Learning framework. The algorithm is verified with simulation data and on an Unmanned Aerial Vehicle (UAV) flying in indoor environment. The loop detection ConvNet alone achieves an accuracy of 98.2% in loop closure detection. With a verification step using the scan matching ConvNet, the false positive rate drops to around 0.001%. The proposed approach processes 6000 pairs of raw LiDAR scans per second on a Nvidia GTX1080 GPU.

It is well known that the rolling shutter effect in images captured with a moving rolling shutter camera causes inaccuracies to 3D reconstructions. The problem is further aggravated with weak visual connectivity from wide baseline images captured with a fast moving camera. In this paper, we propose and implement a pipeline for sparse to dense 3D construction with wide baseline images captured from a fast moving rolling shutter camera. Specifically, we propose a cost function for Bundle Adjustment (BA) that models the rolling shutter effect, incorporates GPS/INS readings, and enforces pairwise smoothness between neighboring poses. We optimize over the 3D structures, camera poses and velocities. We also introduce a novel interpolation scheme for the rolling shutter plane sweep stereo algorithm that allows us to achieve a 7× speed up in the depth map computations for dense reconstruction without losing accuracy. We evaluate our proposed pipeline over a 2.6km image sequence captured with a rolling shutter camera mounted on a moving car.

Automatic welding of tubular TKY joints is an important and challenging task for the marine and offshore industry. In this paper, a framework for tubular joint detection and motion planning is proposed. The pose of the real tubular joint is detected using RGB-D sensors, which is used to obtain a real-to-virtual mapping for positioning the workpiece in a virtual environment. For motion planning, a Bi-directional Transition-based Rapidly exploring Random Tree (BiTRRT) algorithm is used to generate trajectories for reaching the desired goals. The complete framework is verified with experiments, and the results show that the robot welding torch is able to transit without collision to desired goals which are close to the tubular joint.

Automated valet parking services provide great potential to increase the attractiveness of electric vehicles by mitigating their two main current deficiencies: reduced driving ranges and prolonged refueling times. The European research project V-Charge aims at providing this service on designated parking lots using close-to-market sensors only. For this purpose the project developed a prototype capable of performing fully automated navigation in mixed traffic on designated parking lots and GPS-denied parking garages with cameras and ultrasonic sensors only. This paper summarizes the work of the project, comprising advances in network communication and parking space scheduling, multi-camera calibration, semantic mapping concepts, visual localization and motion planning. The project pushed visual localization, environment perception and automated parking to centimetre precision. The developed infrastructure-based camera calibration and semi-supervised semantic mapping concepts greatly reduce maintenance efforts. Results are presented from extensive month-long field tests.

We propose a simple and useful idea based on cross-ratio constraint for wide-baseline matching and 3D reconstruction. Most existing methods exploit feature points and planes from images. Lines have always been considered notorious for both matching and reconstruction due to the lack of good line descriptors. We propose a method to generate and match new points using virtual lines constructed using pairs of keypoints, which are obtained using standard feature point detectors. We use cross-ratio constraints to obtain an initial set of new point matches, which are subsequently used to obtain line correspondences. We develop a method that works for both calibrated and uncalibrated camera configurations. We show compelling line-matching and large-scale 3D reconstruction.

Artefacts that are present in images taken from a moving rolling shutter camera degrade the accuracy of absolute pose estimation. To alleviate this problem, we introduce an addition linear velocity in the camera projection matrix to approximate the motion of the rolling shutter camera. In particular, we derive a minimal solution using the Gröbner Basis that solves for the absolute pose as well as the motion of a rolling shutter camera. We show that the minimal problem requires 5-point correspondences and gives up to 8 real solutions. We also show that our formulation can be extended to use more than 5-point correspondences. We use RANSAC to robustly get all the inliers. In the final step, we relax the linear velocity assumption and do a non-linear refinement on the fuli motion, i.e. linear and angular velocities, and pose of the rolling shutter camera with all the inliers. We verify the feasibility and accuracy of our algorithm with both simulated and real-world datasets.

In this paper, we propose an adaptation of camera projection models for fisheye cameras into the plane-sweeping stereo matching algorithm. This adaptation allows us to do plane-sweeping stereo directly on fisheye images. Our approach also works for other non-pinhole cameras such as omni directional and catadioptric cameras when using the unified projection model. Despite the simplicity of our proposed approach, we are able to obtain full, good quality and high resolution depth maps from the fisheye images. To verify our approach, we show experimental results based on depth maps generated by our approach, and dense models produced from these depth maps.

In this paper, we present our minimal 4-point and linear 8-point algorithms to estimate the relative pose of a multi-camera system with known vertical directions, i.e. known absolute roll and pitch angles. We solve the minimal 4-point algorithm with the hidden variable resultant method and show that it leads to an 8-degree univariate polynomial that gives up to 8 real solutions. We identify a degenerated case from the linear 8-point algorithm when it is solved with the standard Singular Value Decomposition (SVD) method and adopt a simple alternative solution which is easy to implement. We show that our proposed algorithms can be efficiently used within RANSAC for robust estimation. We evaluate the accuracy of our proposed algorithms by comparisons with various existing algorithms for the multi-camera system on simulations and show the feasibility of our proposed algorithms with results from multiple real-world datasets.

The online recalibration of multi-sensor systems is a fundamental problem that must be solved before complex automated systems are deployed in situations such as automated driving. In such situations, accurate knowledge of calibration parameters is critical for the safe operation of automated systems. However, most existing calibration methods for multisensor systems are computationally expensive, use installations of known fiducial patterns, and require expert supervision. We propose an alternative approach called infrastructure-based calibration that is efficient, requires no modification of the infrastructure, and is completely unsupervised. In a survey phase, a computationally expensive simultaneous localization and mapping (SLAM) method is used to build a highly accurate map of a calibration area. Once the map is built, many other vehicles are able to use it for calibration as if it were a known fiducial pattern. We demonstrate the effectiveness of this method to calibrate the extrinsic parameters of a multi-camera system. The method does not assume that the cameras have an overlapping field of view and it does not require an initial guess. As the camera rig moves through the previously mapped area, we match features between each set of synchronized camera images and the map. Subsequently, we find the camera poses and inlier 2D-3D correspondences. From the camera poses, we obtain an initial estimate of the camera extrinsics and rig poses, and optimize these extrinsics and rig poses via non-linear refinement. The calibration code is publicly available as a standalone C++ package.

A potential loop-closure image pair passes the geometric verification test if the number of inliers from the computation of the geometric constraint with RANSAC exceed a pre-defined threshold. The choice of the threshold is critical to the success of identifying the correct loop-closure image pairs. However, the value for this threshold often varies for different datasets and is chosen empirically. In this paper, we propose an unsupervised method that learns the threshold for geometric verification directly from the observed inlier counts of all the potential loop-closure image pairs. We model the distributions of the inlier counts from all the potential loop-closure image pairs with a two components Log-Normal mixture model - one component represents the state of non loop-closure and the other represents the state of loop-closure, and learn the parameters with the Expectation-Maximization algorithm. The intersection of the Log-Normal mixture distributions is the optimal threshold for geometric verification, i.e. the threshold that gives the minimum false positive and negative loop-closures. Our algorithm degenerates when there are too few or no loop-closures and we propose the χ2test to detect this degeneracy. We verify our proposed method with several large-scale datasets collected from both the multi-camera setup and stereo camera.

In this paper, we present a visual ego-motion estimation algorithm for a self-driving car equipped with a close-to-market multi-camera system. By modeling the multi-camera system as a generalized camera and applying the non-holonomic motion constraint of a car, we show that this leads to a novel 2-point minimal solution for the generalized essential matrix where the full relative motion including metric scale can be obtained. We provide the analytical solutions for the general case with at least one inter-camera correspondence and a special case with only intra-camera correspondences. We show that up to a maximum of 6 solutions exist for both cases. We identify the existence of degeneracy when the car undergoes straight motion in the special case with only intra-camera correspondences where the scale becomes unobservable and provide a practical alternative solution. Our formulation can be efficiently implemented within RANSAC for robust estimation. We verify the validity of our assumptions on the motion model by comparing our results on a large real-world dataset collected by a car equipped with 4 cameras with minimal overlapping field-of-views against the GPS/INS ground truth.

In this paper, we model the robust loop-closure pose-graph SLAM problem as a Bayesian network and show that it can be solved with the Classification Expectation-Maximization (EM) algorithm. In particular, we express our robust pose-graph SLAM as a Bayesian network where the robot poses and constraints are latent and observed variables. An additional set of latent variables is introduced as weights for the loop-constraints. We show that the weights can be chosen as the Cauchy function, which are iteratively computed from the errors between the predicted robot poses and observed loop-closure constraints in the Expectation step, and used to weigh the cost functions from the pose-graph loop-closure constraints in the Maximization step. As a result, outlier loop-closure constraints are assigned low weights and exert less influences in the pose-graph optimization within the EM iterations. To prevent the EM algorithm from getting stuck at local minima, we perform the EM algorithm multiple times where the loop constraints with very low weights are removed after each EM process. This is repeated until there are no more changes to the weights. We show proofs of the conceptual similarity between our EM algorithm and the M-Estimator. Specifically, we show that the weight function in our EM algorithm is equivalent to the robust residual function in the M-Estimator. We verify our proposed algorithm with experimental results from multiple simulated and real-world datasets, and comparisons with other existing works.

In this paper, we propose a method to compute the pose-graph loop-closure constraints using multiple non/minimal overlapping field-of-views cameras mounted rigidly on a self-driving car without the need to reconstruct any 3D scene points. In particular, we show that the relative pose with metric scale between two loop-closing pose-graph vertices can be directly obtained from the epipolar geometry of the multicameras system. As a result, we avoid the additional time complexities and uncertainties from the reconstruction of 3D scene points which are needed by standard monocular and stereo approaches. In addition, there is a greater flexibility in choosing a configuration for the multi-camera system to cover a wider field-of-view so as to avoid missing out any loop-closure opportunities. We show that by expressing the point correspondences between two frames as Plücker lines and enforcing the planar motion constraint on the car, we are able to use multiple cameras as one and formulate the relative pose problem for loop-closure as a minimal problem which requires 3-point correspondences that yields up to six real solutions. The RANSAC algorithm is used to determine the correct solution and for robust estimation. We verify our method with results from multiple large-scale real-world data.

We propose an algorithm to estimate the relative camera pose using four feature correspondences and one relative rotation angle measurement. The algorithm can be used for relative pose estimation of a rigid body equipped with a camera and a relative rotation angle sensor which can be either an odometer, an IMU or a GPS/INS system. This algorithm exploits the fact that the relative rotation angles of both the camera and relative rotation angle sensor are the same as the camera and sensor are rigidly mounted to a rigid body. Therefore, knowledge of the extrinsic calibration between the camera and sensor is not required. We carry out a quantitative comparison of our algorithm with the well-known 5-point and 1-point algorithms, and show that our algorithm exhibits the highest level of accuracy.

Future requirements for drastic reduction of CO2production and energy consumption will lead to significant changes in the way we see mobility in the years to come. However, the automotive industry has identified significant barriers to the adoption of electric vehicles, including reduced driving range and greatly increased refueling times. Automated cars have the potential to reduce the environmental impact of driving, and increase the safety of motor vehicle travel. The current state-of-the-art in vehicle automation requires a suite of expensive sensors. While the cost of these sensors is decreasing, integrating them into electric cars will increase the price and represent another barrier to adoption. The V-Charge Project, funded by the European Commission, seeks to address these problems simultaneously by developing an electric automated car, outfitted with close-to-market sensors, which is able to automate valet parking and recharging for integration into a future transportation system. The final goal is the demonstration of a fully operational system including automated navigation and parking. This paper presents an overview of the V-Charge system, from the platform setup to the mapping, perception, and planning sub-systems.

The SFly project is an EU-funded project, with the goal to create a swarm of autonomous vision controlled micro aerial vehicles. The mission in mind is that a swarm of MAV's autonomously maps out an unknown environment, computes optimal surveillance positions and places the MAV's there and then locates radio beacons in this environment. The scope of the work includes contributions on multiple different levels ranging from theoretical foundations to hardware design and embedded programming. One of the contributions is the development of a new MAV, a hexacopter, equipped with enough processing power for onboard computer vision. A major contribution is the development of monocular visual SLAM that runs in real-time onboard of the MAV. The visual SLAM results are fused with IMU measurements and are used to stabilize and control the MAV. This enables autonomous flight of the MAV, without the need of a data link to a ground station. Within this scope novel analytical solutions for fusing IMU and vision measurements have been derived. In addition to the realtime local SLAM, an offline dense mapping process has been developed. For this the MAV's are equipped with a payload of a stereo camera system. The dense environment map is used to compute optimal surveillance positions for a swarm of MAV's. For this an optimiziation technique based on cognitive adaptive optimization has been developed. Finally, the MAV's have been equipped with radio transceivers and a method has been developed to locate radio beacons in the observed environment.

In this paper, we describe our autonomous vision-based quadrotor MAV system which maps and explores unknown environments. All algorithms necessary for autonomous mapping and exploration run on-board the MAV. Using a front-looking stereo camera as the main exteroceptive sensor, our quadrotor achieves these capabilities with both the Vector Field Histogram+ (VFH+) algorithm for local navigation, and the frontier-based exploration algorithm. In addition, we implement the Bug algorithm for autonomous wall-following which could optionally be selected as the substitute exploration algorithm in sparse environments where the frontier-based exploration under-performs. We incrementally build a 3D global occupancy map on-board the MAV. The map is used by the VFH+ and frontier-based exploration in dense environments, and the Bug algorithm for wall-following in sparse environments. During the exploration phase, images from the front-looking camera are transmitted over Wi-Fi to the ground station. These images are input to a large-scale visual SLAM process running off-board on the ground station. SLAM is carried out with pose-graph optimization and loop closure detection using a vocabulary tree. We improve the robustness of the pose estimation by fusing optical flow and visual odometry. Optical flow data is provided by a customized downward-looking camera integrated with a microcontroller while visual odometry measurements are derived from the front-looking stereo camera. We verify our approaches with experimental results.

Bundle adjustment (BA) which produces highly accurate results for visual Simultaneous Localization and Mapping (SLAM) could not be used for Micro-Aerial Vehicles (MAVs) with limited processing power because of its O(N3) complexity. We observed that a consistent ground plane often exists for MAVs flying in both the indoor and outdoor urban environments. Therefore, in this paper, we propose a visual SLAM algorithm that make use of the plane constraint to reduce the complexity of BA. The reduction of complexity is achieved by refining only the current camera pose and most recent map points with BA that minimizes the reprojection errors and perpendicular distances between the most recent map points and the best fit plane with all the pre-existing map points. As a result, our algorithm is approximately constant time since the number of current camera pose and most recent map points remain approximately constant. In addition, the minimization of the perpendicular distances between the plane and map points would enforce consistency between the reconstructed map points and the actual ground plane.

In this paper, we proposed a method to recognize complex human daily activities including body activities and hand gestures simultaneously in an indoor environment. Three wearable motion sensors are attached to the right thigh, the waist, and the right hand of a person, while an optical motion capture system is used to obtain his/her location information. A three-level dynamic Bayesian network is implemented to model the intra-temporal and inter-temporal constraints among the location, body activity and hand gesture. The body activity and hand gesture are estimated using a Bayesian filter and the short-time Viterbi algorithm, which reduces the storage memory and the computational complexity. We conducted experiments in a mock apartment environment and the obtained results showed the effectiveness and accuracy of our algorithms.

In this paper, we present our RS-SLAM algorithm for monocular camera where the proposal distribution is derived from the 5-point RANSAC algorithm and image feature measurement uncertainties instead of using the easily violated constant velocity model. We propose to do another RANSAC sampling within all the inliers that have the best RANSAC score to check for inlier misclassifications in the original correspondences and use all the hypotheses generated from these consensus sets in the proposal distribution. This is to mitigate data association errors (inlier misclassifications) caused by the observation that the consensus set from RANSAC that yields the highest score might not, in practice, contain all the true inliers due to noise on the feature measurements. Hypotheses which are less probable will eventually be eliminated in the particle filter resampling process. We also show in this paper that our monocular approach can be easily extended for stereo camera. Experimental results validate the potential of our approach.

The large collections of datasets for researchers working on the Simultaneous Localization and Mapping problem are mostly collected from sensors such as wheel encoders and laser range finders mounted on ground robots. The recent growing interest in doing visual pose estimation with cameras mounted on micro-aerial vehicles however has made these datasets less useful. In this paper, we describe our work in creating new datasets collected from a sensor suite mounted on a quadrotor platform. Our sensor suite includes a forward looking camera, a downward looking camera, an inertial measurement unit and a Vicon system for groundtruth. We propose the use our datasets as benchmarking tools for future works on visual pose estimation for micro-aerial vehicles. We also show examples of how the datasets could be used for benchmarking visual pose estimation algorithms.