Generating motion sequences conforming to a target style while adhering to the given content prompts requires accommodating both the content and style. In existing methods, the information usually only flows from style to content, which may cause conflict between the style and content, harming the integration. Differently, in this work we build a bidirectional control flow between the style and the content, also adjusting the style towards the content, in which case the style-content collision is alleviated and the dynamics of the style is better preserved in the integration. Moreover, we extend the stylized motion generation from one modality, i.e. the style motion, to multiple modalities including texts and images through contrastive learning, leading to flexible style control on the motion generation. To further boost the performance, we advance the motion diffusion to motion-aligned temporal latent diffusion by developing a novel motion VAE. Extensive experiments demonstrate that our method significantly outperforms previous methods across different datasets, while also enabling multimodal signals control. The code of our method will be made publicly available.

Existing methods for capturing multi-person holistic human motions from a monocular video usually involve integrating the detector, the tracker, and the human pose & shape estimator into a cascaded system. Differently, we develop a one-stage multi-person holistic human motion capture system, which 1) employs only one network, enabling significant benefits from the end-to-end training on a large-scale dataset; 2) enables performance improving of the tracking module during training, avoiding being limited by a pre-trained tracker; 3) captures the motions of all individuals within a single shot, rather than tracking and estimating each person sequentially. In this system, each query within a temporal cross-attention module is responsible for the long motion of a specific individual, implicitly aggregating individual-specific information throughout the entire video. To further boost the proposed system from end-to-end training, we also construct a synthetic human video dataset, with multi-person and whole-body annotations. Extensive experiments across different datasets demonstrate both the efficacy and the efficiency of both the proposed method and the dataset. Codes are avaiable at https://github.com/KenkunLiu/MaQ.

Generating animatable human avatars from a single image is essential for various digital human modeling applications. Existing 3D reconstruction methods often struggle to capture fine details in animatable models, while generative approaches for controllable animation, though avoiding explicit 3D modeling, suffer from viewpoint inconsistencies in extreme poses and computational inefficiencies. In this paper, we address these challenges by leveraging the power of generative models to produce detailed multi-view canonical pose images, which help resolve ambiguities in animatable human reconstruction. We then propose a robust method for 3D reconstruction of inconsistent images, enabling real-time rendering during inference. Specifically, we adapt a transformer-based video generation model to generate multi-view canonical pose images and normal maps, pretraining on a large-scale video dataset to improve generalization. To handle view inconsistencies, we recast the reconstruction problem as a 4D task and introduce an efficient 3D modeling approach using 4D Gaussian Splatting. Experiments demonstrate that our method achieves photorealistic, real-time animation of 3D human avatars from in-the-wild images, showcasing its effectiveness and generalization capability. Our code will be available on https://github.com/aigc3d/AniGS.

Language plays a vital role in the realm of human motion. Existing methods have largely depended on CLIP text embeddings for motion generation, yet they fall short in effectively aligning language and motion due to CLIP’s pretraining on static image-text pairs. This work introduces LaMP, a novel Language-Motion Pretraining model, which transitions from a language-vision to a more suitable language-motion latent space. It addresses key limitations by generating motion-informative text embeddings, significantly enhancing the relevance and semantics of generated motion sequences. With LaMP, we advance three key tasks: text-to-motion generation, motion-text retrieval, and motion captioning through aligned language-motion representation learning. For generation, LaMP instead of CLIP provides the text condition, and an autoregressive masked prediction is designed to achieve mask modeling without rank collapse in transformers. For retrieval, motion features from LaMP’s motion transformer interact with query tokens to retrieve text features from the text transformer, and vice versa. For captioning, we finetune a large language model with the language-informative motion features to develop a strong motion captioning model. In addition, we introduce the LaMP-BertScore metric to assess the alignment of generated motions with textual descriptions. Extensive experimental results on multiple datasets demonstrate substantial improvements over previous methods across all three tasks. Project page: https://aigc3d.github.io/LaMP

Tri-plane-like representations have been widely adopted in 3D-aware GANs for head image synthesis and other 3D object/scene modeling tasks due to their efficiency. However, querying features via Cartesian coordinate projection often leads to feature entanglement, which results in mirroring artifacts. A recent work, SphereHead, attempted to address this issue by introducing spherical tri-planes based on a spherical coordinate system. While it successfully mitigates feature entanglement, SphereHead suffers from uneven mapping between the square feature maps and the spherical planes, leading to inefficient feature map utilization during rendering and difficulties in generating fine image details.Moreover, both tri-plane and spherical tri-plane representations share a subtle yet persistent issue: feature penetration across convolutional channels can cause interference between planes, particularly when one plane dominates the others (see Fig. 1). These challenges collectively prevent tri-plane-based methods from reaching their full potential. In this paper, we systematically analyze these problems for the first time and propose innovative solutions to address them. Specifically, we introduce a novel hybrid-plane (hy-plane for short) representation that combines the strengths of both planar and spherical planes while avoiding their respective drawbacks. We further enhance the spherical plane by replacing the conventional theta-phi warping with a novel near-equal-area warping strategy, which maximizes the effective utilization of the square feature map. In addition, our generator synthesizes a single-channel unified feature map instead of multiple feature maps in separate channels, thereby effectively eliminating feature penetration. With a series of technical improvements, our hy-plane representation enables our method, HyPlaneHead, to achieve state-of-the-art performance in full-head image synthesis.

Generating highly dynamic and photorealistic portrait animations driven by audio and skeletal motion remains challenging due to the need for precise lip synchronization, natural facial expressions, and high-fidelity body motion dynamics. We propose a human-preference-aligned diffusion framework that addresses these challenges through two key innovations. First, we introduce direct preference optimization tailored for human-centric animation, leveraging a curated dataset of human preferences to align generated outputs with perceptual metrics for portrait motion-video alignment and naturalness of expression. Second, the proposed temporal motion modulation resolves spatiotemporal resolution mismatches by reshaping motion conditions into dimensionally aligned latent features through temporal channel redistribution and proportional feature expansion, preserving the fidelity of high-frequency motion details in diffusion-based synthesis. The proposed mechanism is complementary to existing UNet and DiT-based portrait diffusion approaches, and experiments demonstrate obvious improvements in lip-audio synchronization, expression vividness, body motion coherence over baseline methods, alongside notable gains in human preference metrics. Code and data for this paper are at https://github.com/fudan-generative-vision/hallo4.

Panoramic images enable deeper understanding and more holistic perception of 360° surrounding environment, which can naturally encode enriched scene context information compared to standard perspective image. Previous work has made lots of effort to solve the scene understanding task in a hybrid solution based on 2D-3D geometric reasoning, thus each sub-task is processed separately and few correlations are explored in this procedure. In this paper, we propose a fully 3D method for holistic indoor scene understanding which recovers the objects' shapes, oriented bounding boxes and the 3D room layout simultaneously from a single panorama. To maximize the exploration of the rich context information, we design a transformer-based context module to predict the representation and relationship among each component of the scene. In addition, we introduce a new dataset for scene understanding, including photo-realistic panoramas, high-fidelity depth images, accurately annotated room layouts, oriented object bounding boxes and shapes. Experiments on the synthetic and new datasets demonstrate that our method outperforms previous panoramic scene understanding methods in terms of both layout estimation and 3D object detection.

State-of-the-art man-made shape generative models usually adopt established generative models under a suitable implicit shape representation. A common theme is to perform distribution alignment, which does not explicitly model important shape priors. As a result, many synthetic shapes are not connected. Other synthetic shapes present problems of physical stability and geometric feasibility. This paper introduces a novel latent diffusion shape-generative model regularized by a quality checker that outputs a score of a latent code. The scoring function employs a learned function that provides a geometric feasibility score and a deterministic procedure to quantify a physical stability score. The key to our approach is a new diffusion procedure that combines the discrete empirical data distribution and a continuous distribution induced by the quality checker. We introduce a principled approach to determine the trade-off parameters for learning the denoising network at different noise levels. Experimental results show that our approach outperforms state-of-the-art shape generations quantitatively and qualitatively on ShapeNet-v2.

Lifting 2D diffusion for 3D generation is a challenging problem due to the lack of geometric prior and the complex entanglement of materials and lighting in natural images. Existing methods have shown promise by first creating the geometry through score-distillation sampling (SDS) applied to rendered surface normals, followed by appearance modeling. However, relying on a 2D RGB diffusion model to optimize surface normals is suboptimal due to the distribution discrepancy between natural images and normals maps, leading to instability in optimization. In this paper, recognizing that the normal and depth information effectively describe scene geometry and be auto-matically estimated from images, we propose to learn a generalizable Normal-Depth diffusion model for 3D generation. We achieve this by training on the large-scale LAION dataset together with the generalizable image-to-depth and normal prior models. In an attempt to alleviate the mixed illumination effects in the generated materials, we introduce an albedo diffusion model to impose data-driven constraints on the albedo component. Our experiments show that when integrated into existing text-to-3D pipelines, our models significantly enhance the detail richness, achieving state-of-the-art results. Our project page is at https://aigc3d.github.io/richdreamer/.

Generalizable 3D object reconstructionfrom single-view RGB-D images remains a challenging task, particularly with real-world data. Current state-of-the-art methods develop Transformer-based implicit field learning, necessitating an intensive learning paradigm that requires dense query-supervision uniformly sampled throughout the entire space. We propose a novel approach, IPoD, which harmonizes implicit field learning with point diffusion. This approach treats the query points for implicit field learning as a noisy point cloud for iterative denoising, allowing for their dynamic adaptation to the target object shape. Such adaptive query points harness diffusion learning's capability for coarse shape recovery and also enhances the implicit representation's ability to delineate finer details. Besides, an additional self-conditioning mechanism is designed to use implicit predictions as the guidance of diffusion learning, leading to a cooperative system. Experiments conducted on the CO3D-v2 dataset affirm the superiority of IPoD, achieving 7.8% improvement in F-score and 28.6% in Chamfer distance over existing methods. The generalizability of IPoD is also demonstrated on the MVImgNet dataset. Our project page is at https://yushuang-wu.github.io/IPoD.

Motion generation from discrete quantization offers many advantages over continuous regression, but at the cost of inevitable approximation errors. Previous methods usually quantize the entire body pose into one code, which not only faces the difficulty in encoding all joints within one vector but also loses the spatial relationship between different joints. Differently, in this work we quantize each individual joint into one vector, which i) simplifies the quantization process as the complexity associated with a single joint is markedly lower than that of the entire pose; ii) maintains a spatial-temporal structure that preserves both the spatial relationships among joints and the temporal movement patterns; iii) yields a 2D token map, which enables the application of various 2D operations widely used in 2D images. Grounded in the 2D motion quantization, we build a spatial-temporal modeling framework, where 2D joint VQVAE, temporal-spatial 2D masking technique, and spatial-temporal 2D attention are proposed to take advantage of spatial-temporal signals among the 2D tokens. Extensive experiments demonstrate that our method significantly outperforms previous methods across different datasets, with a $26.6\%$ decrease of FID on HumanML3D and a $29.9\%$ decrease on KIT-ML.

This paper studies the problem of estimating physical properties (system identification) through visual observations. To facilitate geometry-aware guidance in physical property estimation, we introduce a novel hybrid framework that leverages 3D Gaussian representation to not only capture explicit shapes but also enable the simulated continuum to render object masks as 2D shape surrogates during training. We propose a new dynamic 3D Gaussian framework based on motion factorization to recover the object as 3D Gaussian point sets across different time states. Furthermore, we develop a coarse-to-fine filling strategy to generate the density fields of the object from the Gaussian reconstruction, allowing for the extraction of object continuums along with their surfaces and the integration of Gaussian attributes into these continuum. In addition to the extracted object surfaces, the Gaussian-informed continuum also enables the rendering of object masks during simulations, serving as 2D-shape guidance for physical property estimation. Extensive experimental evaluations demonstrate that our pipeline achieves state-of-the-art performance across multiple benchmarks and metrics. Additionally, we illustrate the effectiveness of the proposed method through real-world demonstrations, showcasing its practical utility. Our project page is at https://jukgei.github.io/project/gic.

Novel view synthesis aims at rendering any posed images from sparse observations of the scene. Recently, neural radiance fields (NeRF) have demonstrated their effectiveness in synthesizing novel views of a bounded scene. However, most existing methods cannot be directly extended to 360° unbounded scenes where the camera orientations and scene depths are unconstrained with large variations. In this paper, we present a spherical radiance field (SRF) for efficient novel view synthesis in 360° unbounded scenes. Specifically, we represent a 3D scene as multiple concentric spheres with different radii. In particular, each sphere encodes its corresponding layered scene into implicit representations and is parameterized with an equirectangular projection image. A shallow multi-layer perceptron (MLP) is then used to infer the density and color from these sphere representations for volume rendering. Moreover, an occupancy grid is introduced to cache the density field and guide the ray sampling, which accelerates the training and rendering procedures by reducing the number of samples along the ray. Experiments show that our method can well fit 360° unbounded scenes and produces state-of-the-art results on three benchmark datasets with less than 30 minutes of training time on a 3090 GPU, surpassing Mip-NeRF 360 with a 400× speedup. In addition, our method achieves competitive performance in terms of both accuracy and efficiency on a bounded dataset. Project page: https://minglin-chen.github.io/SphericalRF.

MVImgNet is a large-scale dataset that contains multi-view images of ~220k real-world objects in 238 classes. As a counterpart of ImageNet, it introduces 3D visual signals via multi-view shooting, making a soft bridge between 2D and 3D vision. This paper constructs the MVImgNet2.0 dataset that expands MVImgNet into a total of ~520k objects and 515 categories, which derives a 3D dataset with a larger scale that is more comparable to ones in the 2D domain. In addition to the expanded dataset scale and category range, MVImgNet2.0 is of a higher quality than MVImgNet owing to four new features: (i) most shoots capture 360° views of the objects, which can support the learning of object reconstruction with completeness; (ii) the segmentation manner is advanced to produce foreground object masks of higher accuracy; (iii) a more powerful structure-from-motion method is adopted to derive the camera pose for each frame of a lower estimation error; (iv) higher-quality dense point clouds are reconstructed via advanced methods for objects captured in 360 ° views, which can serve for downstream applications. Extensive experiments confirm the value of the proposed MVImgNet2.0 in boosting the performance of large 3D reconstruction models. MVImgNet2.0 will be public at luyues.github.io/mvimgnet2 , including multi-view images of all 520k objects, the reconstructed high-quality point clouds, and data annotation codes, hoping to inspire the broader vision community.

This work addresses the challenge of high-quality surface normal estimation from monocular colored inputs (i.e., images and videos), a field which has recently been revolutionized by repurposing diffusion priors. However, previous attempts still struggle with stochastic inference, conflicting with the deterministic nature of the Image2Normal task, and costly ensembling step, which slows down the estimation process. Our method, StableNormal, mitigates the stochasticity of the diffusion process by reducing inference variance, thus producing "Stable-and-Sharp" normal estimates without any additional ensembling process. StableNormal works robustly under challenging imaging conditions, such as extreme lighting, blurring, and low quality. It is also robust against transparent and reflective surfaces, as well as cluttered scenes with numerous objects. Specifically, StableNormal employs a coarse-to-fine strategy, which starts with a one-step normal estimator (YOSO) to derive an initial normal guess, that is relatively coarse but reliable, then followed by a semantic-guided refinement process (SG-DRN) that refines the normals to recover geometric details. The effectiveness of StableNormal is demonstrated through competitive performance in standard datasets such as DIODE-indoor, iBims, ScannetV2 and NYUv2, and also in various downstream tasks, such as surface reconstruction and normal enhancement. These results evidence that StableNormal retains both the "stability" and "sharpness" for accurate normal estimation. StableNormal represents a baby attempt to repurpose diffusion priors for deterministic estimation. To democratize this, code and models have been publicly available in hf.co/Stable-X.

Recently, Neural Radiance Fields (NeRF) has exhibited significant success in novel view synthesis, surface reconstruction, etc. However, since no physical reflection is considered in its rendering pipeline, NeRF mistakes the reflection in the mirror as a separate virtual scene, leading to the inaccurate reconstruction of the mirror and multi-view inconsistent reflections in the mirror. In this paper, we present a novel neural rendering framework, named Mirror-NeRF, which is able to learn accurate geometry and reflection of the mirror and support various scene manipulation applications with mirrors, such as adding new objects or mirrors into the scene and synthesizing the reflections of these new objects in mirrors, controlling mirror roughness, etc. To achieve this goal, we propose a unified radiance field by introducing the reflection probability and tracing rays following the light transport model of Whitted Ray Tracing, and also develop several techniques to facilitate the learning process. Experiments and comparisons on both synthetic and real datasets demonstrate the superiority of our method. The code and supplementary material are available on the project webpage: https://zju3dv.github.io/Mirror-NeRF/.

The integration of multiple cameras and 3D Li-DARs has become basic configuration of augmented reality devices, robotics, and autonomous vehicles. The calibration of multi-modal sensors is crucial for a system to properly function, but it remains tedious and impractical for mass production. Moreover, most devices require re-calibration after usage for certain period of time. In this paper, we propose a single-shot solution for calibrating extrinsic transformations among multiple cameras and 3D LiDARs. We establish a panoramic infrastructure, in which a camera or LiDAR can be robustly localized using data from single frame. Experiments are conducted on three devices with different camera-LiDAR configurations, showing that our approach achieved comparable calibration accuracy with the state-of-the-art approaches but with much greater efficiency.

Structure-from-motion (SfM) largely relies on feature tracking. In image sequences, if disjointed tracks caused by objects moving in and out of the field of view, occasional occlusion, or image noise are not handled well, corresponding SfM could be affected. This problem becomes severer for large-scale scenes, which typically requires to capture multiple sequences to cover the whole scene. In this paper, we propose an efficient non-consecutive feature tracking framework to match interrupted tracks distributed in different subsequences or even in different videos. Our framework consists of steps of solving the feature "dropout" problem when indistinctive structures, noise or large image distortion exists, and of rapidly recognizing and joining common features located in different subsequences. In addition, we contribute an effective segment-based coarse-to-fine SfM algorithm for robustly handling large data sets. Experimental results on challenging video data demonstrate the effectiveness of the proposed system.

We present a novel real-time monocular SLAM system which can robustly work in dynamic environments. Different to the traditional methods, our system allows parts of the scene to be dynamic or the whole scene to gradually change. The key contribution is that we propose a novel online keyframe representation and updating method to adaptively model the dynamic environments, where the appearance or structure changes can be effectively detected and handled. We reliably detect the changed features by projecting them from the keyframes to current frame for appearance and structure comparison. The appearance change due to occlusions also can be reliably detected and handled. The keyframes with large changed areas will be replaced by newly selected frames. In addition, we propose a novel prior-based adaptive RANSAC algorithm (PARSAC) to efficiently remove outliers even when the inlier ratio is rather low, so that the camera pose can be reliably estimated even in very challenging situations. Experimental results demonstrate that the proposed system can robustly work in dynamic environments and outperforms the state-of-the-art SLAM systems (e.g. PTAM).

We present a novel keyframe selection and recognition method for robust markerless real-time camera tracking. Our system contains an offline module to select features from a group of reference images and an online module to match them to the input live video in order to quickly estimate the camera pose. The main contribution lies in constructing an optimal set of keyframes from the input reference images, which are required to approximately cover the entire space and at the same time minimize the content redundancy amongst the selected frames. This strategy not only greatly saves the computation, but also helps significantly reduce the number of repeated features so as to improve the camera tracking quality. Our system also employs a parallel-computing scheme with multi-CPU hardware architecture. Experimental results show that our method dramatically enhances the computation efficiency and eliminates the jittering artifacts.

Abstract High‐quality video editing usually requires accurate layer separation in order to resolve occlusions. However, most of the existing bilayer segmentation algorithms require either considerable user intervention or a simple stationary camera configuration with known background, which is difficult to meet for many real world online applications. This paper demonstrates that various visually appealing montage effects can be online created from a live video captured by a rotating camera, by accurately retrieving the camera state and segmenting out the dynamic foreground. The key contribution is that a novel fast bilayer segmentation method is proposed which can effectively extract the dynamic foreground under rotational camera configuration, and is robust to imperfect background estimation and complex background colors. Our system can create a variety of live visual effects, including but not limited to, realistic virtual object insertion, background substitution and blurring, non‐photorealistic rendering and camouflage effect. A variety of challenging examples demonstrate the effectiveness of our method.

Compared to still image editing, content-based video editing faces the additional challenges of maintaining the spatiotemporal consistency with respect to geometry. This brings up difficulties of seamlessly modifying video content, for instance, inserting or removing an object. In this paper, we present a new video editing system for creating spatiotemporally consistent and visually appealing refilming effects. Unlike the typical filming practice, our system requires no labor-intensive construction of 3D models/surfaces mimicking the real scene. Instead, it is based on an unsupervised inference of view-dependent depth maps for all video frames. We provide interactive tools requiring only a small amount of user input to perform elementary video content editing, such as separating video layers, completing background scene, and extracting moving objects. These tools can be utilized to produce a variety of visual effects in our system, including but not limited to video composition, "predator" effect, bullet-time, depth-of-field, and fog synthesis. Some of the effects can be achieved in real time.