Transparent image layer generation plays a significant role in digital art and design workflows. Existing methods typically decompose transparent layers from a single RGB image using a set of tools or generate multiple transparent layers sequentially. Despite some promising results, these methods often limit their ability to model global layout, physically plausible interactions, and visual effects such as shadows and reflections with high alpha quality due to limited shared global context among layers. To address this issue, we propose PSDiffusion, a unified diffusion framework that leverages image composition priors from pre-trained image diffusion model for simultaneous multi-layer text-to-image generation. Specifically, our method introduces a global layer interaction mechanism to generate layered images collaboratively, ensuring both individual layer quality and coherent spatial and visual relationships across layers. We include extensive experiments on benchmark datasets to demonstrate that PSDiffusion is able to outperform existing methods in generating multi-layer images with plausible structure and enhanced visual fidelity.

Abstract Existing image restoration methods primarily rely on the posterior distribution of natural images but are often limited by their dependence on known degradations and supervised training. To this end, we propose Generative Diffusion Prior (GDP), an unsupervised sampling-based framework that effectively models posterior distributions for image and video restoration. GDP utilizes a single pre-trained denoising diffusion probabilistic model (DDPM) to solve a wide range of linear, non-linear, and blind inverse problems without explicit degradation assumptions. Specifically, GDP systematically explores a conditional guidance protocol, which proves more practical and effective than conventional methods of adding guidance. Furthermore, GDP incorporates a degradation model optimization mechanism during the denoising process, enabling blind image restoration. Besides, we introduce a patch-based strategy, allowing GDP to handle images of arbitrary resolution. We extensively evaluate GDP on multiple image and video restoration tasks, including super-resolution, deblurring, inpainting, and colorization, as well as more challenging applications such as low-light enhancement, HDR recovery, and LDR video enhancement. Experimental results demonstrate that GDP outperforms leading unsupervised methods across diverse benchmarks in both reconstruction accuracy and perceptual quality, while demonstrating robust generalization to images and videos of any size. Our project page at https://generativediffusionprior.github.io/.

The field of neural rendering has seen remarkable progress, driven by advancements in generative models and differentiable rendering techniques. While 2D diffusion has achieved notable success, the development of a unified 3D diffusion pipeline remains an open challenge. This paper presents a novel framework, LN3Diff++, designed to bridge this gap and facilitate fast, high-quality, and versatile conditional 3D generation. Our method leverages a 3D-aware architecture and a variational autoencoder (VAE) to encode input image(s) into a structured, compact 3D latent space. The latent representation is then decoded by a transformer-based decoder into a high-capacity 3D neural field. By training a diffusion model on this 3D-aware latent space, our method achieves superior performance for category-specific 3D generation on ShapeNet and FFHQ, as well as category-free image/text-conditioned 3D generation over Objaverse. Moreover, it surpasses existing 3D diffusion methods in inference speed, requiring no per-instance optimization.

Video inpainting, which aims to fill missing regions with visually coherent content, has emerged as a crucial technique for creative applications such as editing. While existing approaches achieve visual consistency or text-guided generation, they often struggle to balance coherence and creative diversity. In this work, we introduce VideoRepainter, a two-stage framework that allows users to inpaint a keyframe using established image-level techniques, then propagate the changes to other frames. Our approach can leverage state-of-the-art image models for keyframe manipulation, thereby easing the burden of the video-inpainting process. To this end, we integrate an image-to-video model with a symmetric condition mechanism to address ambiguity caused by direct mask downsampling. We further explore efficient strategies for mask synthesis and parameter tuning to reduce costs in data processing and model training. Evaluations demonstrate our method achieves superior results in both visual fidelity and content diversity compared to existing approaches, providing a practical solution for creative video manipulation. See our Project Page for more details.

Recent advances in 3D Gaussian Splatting (3DGS) have demonstrated significant potential over traditional rendering techniques, attracting widespread attention from both industry and academia. However, real-time rendering with 3DGS remains a challenging problem, particularly in large-scale, high-resolution scenes due to the presence of numerous anisotropic Gaussian representations, and it has not been extensively explored. To address this challenge, we introduce FlashGS, an open-source CUDA library with Python bindings, featuring comprehensive algorithm design and optimizations, including redundancy elimination, adaptive scheduling, and efficient pipelining. First, we eliminate substantial redundant computations through precise Gaussian intersection tests, leveraging the intrinsic mechanism of the 3DGS rasterizer. During task partitioning, we propose an adaptive scheduling strategy that accounts for variations in Gaussian size and shape. Additionally, we design a multi-stage pipelining strategy for color computation in the rendering process, further accelerating performance. We conduct an extensive evaluation of FlashGS across a diverse range of synthetic and real-world 3D scenes, encompassing scene sizes of up to 2.7 km2cityscape and resolutions of up to over 4K. Our approach improves 3DGS rendering performance by an order of magnitude, achieving an average speedup of 7.2×, and rendering at a minimum of 125.9 FPS, setting a new state-of-the-art in real-time 3DGS rendering. https://github.com/InternLandMark/FlashGS.

Seamless integration of both aerial and street view images remains a significant challenge in neural scene reconstruction and rendering. Existing methods predominantly focus on single domain, limiting their applications in immersive environments, which demand extensive free view exploration with large view changes both horizontally and vertically. We introduce Horizon-Gs, a novel approach built upon Gaussian Splatting techniques, tackles the unified reconstruction and rendering for aerial and street views. Our method addresses the key challenges of combining these perspectives with a new training strategy, overcoming viewpoint discrepancies to generate high-fidelity scenes. We also curate a high-quality aerial-to-ground views dataset encompassing both synthetic and real-world scene to advance further research. Experiments across diverse urban scene datasets confirm the effectiveness of our method.

The scaling law has been validated in various domains, such as natural language processing (NLP) and massive computer vision tasks; however, its application to motion generation remains largely unexplored. In this paper, we introduce a scalable motion generation framework that includes the motion tokenizer Motion FSQ-VAE and a text-prefix autoregressive transformer. Through comprehensive experiments, we observe the scaling behavior of this system. For the first time, we confirm the existence of scaling laws within the context of motion generation. Specifically, our results demonstrate that the normalized test loss of our prefix autoregressive models adheres to a logarithmic law in relation to compute budgets. Furthermore, we also confirm the power law between Non-Vocabulary Parameters, Vocabulary Parameters, and Data Tokens with respect to compute budgets respectively. Leveraging the scaling law, we predict the optimal transformer size, vocabulary size, and data requirements for a compute budget of 1e18. The test loss of the system, when trained with the optimal model size, vocabulary size, and required data, aligns precisely with the predicted test loss, thereby validating the scaling law. Project page: https://shunlinlu.github.io/ScaMo/

Synthesizing diverse and physically plausible Human-Scene Interactions (HSI) is pivotal for both computer animation and embodied AI. Despite encouraging progress, current methods mainly focus on developing separate controllers, each specialized for a specific interaction task. This significantly hinders the ability to tackle a wide variety of challenging HSI tasks that require the integration of multiple skills, e.g. sitting down while carrying an object (see Fig. 1). To address this issue, we present TokenHSI, a single, unified transformer-based policy capable of multi-skill unification and flexible adaptation. The key insight is to model the humanoid proprioception as a separate shared token and combine it with distinct task tokens via a masking mechanism. Such a unified policy enables effective knowledge sharing across skills, thereby facilitating the multi-task training. Moreover, our policy architecture supports variable length inputs, enabling flexible adaptation of learned skills to new scenarios. By training additional task tokenizers, we can not only modify the geometries of interaction targets but also coordinate multiple skills to address complex tasks. The experiments demonstrate that our approach can significantly improve versatility, adaptability, and extensibility in various HSI tasks.

Recent advances in generative models have enabled high-quality 3D character reconstruction from multi-modal. However, animating these generated characters remains a challenging task, especially for complex elements like garments and hair, due to the lack of large-scale datasets and effective rigging methods. To address this gap, we curate AnimeRig, a large-scale dataset with detailed skeleton and skinning annotations. Building upon this, we propose DRiVE, a novel framework for generating and rigging 3D human characters with intricate structures. Unlike existing methods, DRiVE utilizes a 3D Gaussian representation, facilitating efficient animation and high-quality rendering. We further introduce GSDiff, a 3D Gaussian-based diffusion module that predicts joint positions as spatial distributions, overcoming the limitations of regression-based approaches. Extensive experiments demonstrate that DRiVE achieves precise rigging results, enabling realistic dynamics for clothing and hair, and surpassing previous methods in both quality and versatility. Code, dataset and visualization results are available at https://DRiVEAvatar.github.io/.

On top of Segment Anything Model (SAM), SAM 2 further extends its capability from image to video inputs through a memory bank mechanism and obtains a remarkable performance compared with previous methods, making it a foundation model for video segmentation task. In this paper, we aim at making SAM 2 much more efficient so that it even runs on mobile devices while maintaining a comparable performance. Despite several works optimizing SAM for better efficiency, we find they are not sufficient for SAM 2 because they all focus on compressing the image encoder, while our benchmark shows that the newly introduced memory attention blocks are also the latency bottleneck. Given this observation, we propose EdgeTAM, which leverages a novel 2D Spatial Perceiver to reduce the computational cost. In particular, the proposed 2D Spatial Perceiver encodes the densely stored frame-level memories with a lightweight Transformer that contains a fixed set of learnable queries. Given that video segmentation is a dense prediction task, we find preserving the spatial structure of the memories is essential so that the queries are split into global-level and patch-level groups. We also propose a distillation pipeline that further improves the performance without inference overhead. As a result, EdgeTAM achieves 87.7, 70.0, 72.3, and 71.7 ${\mathcal{J}}{{\& }}{\mathcal{F}}$ on DAVIS 2017, MOSE, SA-V val, and SA-V test, while running at 16 FPS on iPhone 15 Pro Max. The code and models are available here.

We present a unified and generalizable framework for synthesizing view-consistent and temporally coherent avatars from a single image, addressing the challenging task of single-image avatar generation. Existing diffusion-based methods often condition on sparse human templates (e.g., depth or normal maps), which leads to multi-view and temporal inconsistencies due to the mismatch between these signals and the true appearance of the subject. Our approach bridges this gap by combining the reconstruction power of regression-based 3D human reconstruction with the generative capabilities of a diffusion model. In a first step, an initial 3D reconstructed human through a generalized NeRF provides comprehensive conditioning, ensuring high-quality synthesis faithful to the reference appearance and structure. Subsequently, the derived geometry and appearance from the generalized NeRF serve as input to a video-based diffusion model. This strategic integration is pivotal for enforcing both multi-view and temporal consistency throughout the avatar's generation. Empirical results underscore the superior generalization ability of our proposed method, demonstrating its effectiveness across diverse in-domain and out-of-domain in-the-wild datasets.

Generating human videos from a single image while ensuring high visual quality and precise control is a challenging task, especially in complex scenarios involving multiple individuals and interactions with objects. Existing methods, while effective for single-human cases, often fail to handle the intricacies of multi-identity interactions because they struggle to associate the correct pairs of human appearance and pose condition and model the distribution of 3D-aware dynamics. To address these limitations, we present \emph{Structural Video Diffusion}, a novel framework designed for generating realistic multi-human videos. Our approach introduces two core innovations: identity-specific embeddings to maintain consistent appearances across individuals and a structural learning mechanism that incorporates depth and surface-normal cues to model human-object interactions. Additionally, we expand existing human video dataset with 25K new videos featuring diverse multi-human and object interaction scenarios, providing a robust foundation for training. Experimental results demonstrate that Structural Video Diffusion achieves superior performance in generating lifelike, coherent videos for multiple subjects with dynamic and rich interactions, advancing the state of human-centric video generation. Code is available at https://github.com/zhenzhiwang/Multi-HumanVid

Controllability plays a crucial role in video generation, as it allows users to create and edit content more precisely. Existing models, however, lack control of camera pose that serves as a cinematic language to express deeper narrative nuances. To alleviate this issue, we introduce \method, enabling accurate camera pose control for video diffusion models. Our approach explores effective camera trajectory parameterization along with a plug-and-play camera pose control module that is trained on top of a video diffusion model, leaving other modules of the base model untouched. Moreover, a comprehensive study on the effect of various training datasets is conducted, suggesting that videos with diverse camera distributions and similar appearance to the base model indeed enhance controllability and generalization. Experimental results demonstrate the effectiveness of \method in achieving precise camera control with different video generation models, marking a step forward in the pursuit of dynamic and customized video storytelling from textual and camera pose inputs.

Recent advancements in diffusion models and large-scale datasets have revolutionized image and video generation, with increasing focus on 3D content generation. While existing methods show promise, they face challenges in input formats, latent space structures, and output representations. This paper introduces a novel 3D generation framework that addresses these issues, enabling scalable and high-quality 3D generation with an interactive Point Cloud-structured Latent space. Our approach utilizes a VAE with multi-view posed RGB-D-N renderings as input, features a unique latent space design that preserves 3D shape information, and incorporates a cascaded latent flow-based model for improved shape-texture disentanglement. The proposed method, GaussianAnything, supports multi-modal conditional 3D generation, allowing for point cloud, caption, and single-view image inputs. Experimental results demonstrate superior performance on various datasets, advancing the state-of-the-art in 3D content generation.

Reconstructing 3D objects into editable programs is pivotal for applications like reverse engineering and shape editing. However, existing methods often rely on limited domain-specific languages (DSLs) and small-scale datasets, restricting their ability to model complex geometries and structures. To address these challenges, we introduce MeshLLM, a novel framework that reconstructs complex 3D objects from point clouds into editable Blender Python scripts. We develop a comprehensive set of expressive Blender Python APIs capable of synthesizing intricate geometries. Leveraging these APIs, we construct a large-scale paired object-code dataset, where the code for each object is decomposed into distinct semantic parts. Subsequently, we train a multimodal large language model (LLM) that translates 3D point cloud into executable Blender Python scripts. Our approach not only achieves superior performance in shape-to-code reconstruction tasks but also facilitates intuitive geometric and topological editing through convenient code modifications. Furthermore, our code-based representation enhances the reasoning capabilities of LLMs in 3D shape understanding tasks. Together, these contributions establish MeshLLM as a powerful and flexible solution for programmatic 3D shape reconstruction and understanding.

Realistic 3D indoor scene synthesis is vital for embodied AI and digital content creation. It can be naturally divided into two subtasks: object generation and layout generation. While recent generative models have significantly advanced object-level quality and controllability, layout generation remains challenging due to limited datasets. Existing methods either overfit to these datasets or rely on predefined constraints to optimize numerical layout that sacrifice flexibility. As a result, they fail to generate scenes that are both open-vocabulary and aligned with fine-grained user instructions. We introduce DirectLayout, a framework that directly generates numerical 3D layouts from text descriptions using generalizable spatial reasoning of large language models (LLMs). DirectLayout decomposes the generation into three stages: producing a Bird's-Eye View (BEV) layout, lifting it into 3D space, and refining object placements. To enable explicit spatial reasoning and help the model grasp basic principles of object placement, we employ Chain-of-Thought (CoT) Activation based on the 3D-Front dataset. Additionally, we design CoT-Grounded Generative Layout Reward to enhance generalization and spatial planning. During inference, DirectLayout addresses asset-layout mismatches via Iterative Asset-Layout Alignment through in-context learning. Extensive experiments demonstrate that DirectLayout achieves impressive semantic consistency, generalization and physical plausibility.

Object compositing offers significant promise for augmented reality (AR) and embodied intelligence applications. Existing approaches predominantly focus on single-image scenarios or intrinsic decomposition techniques, facing challenges with multi-view consistency, complex scenes, and diverse lighting conditions. Recent inverse rendering advancements, such as 3D Gaussian and diffusion-based methods, have enhanced consistency but are limited by scalability, heavy data requirements, or prolonged reconstruction time per scene. To broaden its applicability, we introduce MV-CoLight, a two-stage framework for illumination-consistent object compositing in both 2D images and 3D scenes. Our novel feed-forward architecture models lighting and shadows directly, avoiding the iterative biases of diffusion-based methods. We employ a Hilbert curve-based mapping to align 2D image inputs with 3D Gaussian scene representations seamlessly. To facilitate training and evaluation, we further introduce a large-scale 3D compositing dataset. Experiments demonstrate state-of-the-art harmonized results across standard benchmarks and our dataset, as well as casually captured real-world scenes demonstrate the framework's robustness and wide generalization.

Vision Transformers (ViTs) have emerged as the dominant architecture for visual processing tasks, demonstrating excellent scalability with increased training data and model size. However, recent work has identified the emergence of artifact tokens in ViTs that are incongruous with local semantics. These anomalous tokens degrade ViT performance in tasks that require fine-grained localization or structural coherence. An effective mitigation of this issue is the addition of register tokens to ViTs, which implicitly ''absorb'' the artifact term during training. Given the availability of existing large-scale pre-trained ViTs, in this paper we seek to add register tokens to existing models without retraining the models from scratch, which is infeasible considering their size. Specifically, we propose Post Hoc Registers (**PH-Reg**), an efficient self-distillation method that integrates registers into an existing ViT without requiring additional labeled data and full retraining. PH-Reg initializes both teacher and student networks from the same pre-trained ViT. The teacher remains frozen and unmodified, while the student is augmented with randomly initialized register tokens. By applying test-time augmentation to the teacher’s inputs, we generate denoised dense embeddings free of artifacts, which are then used to optimize only a small subset of unlocked student weights. We show that our approach can effectively reduce the number of artifact tokens, improving the segmentation and depth prediction of the student ViT under zero-shot and linear probing.

The advancement of Embodied AI heavily relies on large-scale, simulatable 3D scene datasets characterized by scene diversity and realistic layouts.However, existing datasets typically suffer from limitations in data scale or diversity, sanitized layouts lacking small items, and severe object collisions.To address these shortcomings, we introduce \textbf{InternScenes}, a novel large-scale simulatable indoor scene dataset comprising approximately 40,000 diverse scenes by integrating three disparate scene sources, \ie, real-world scans, procedurally generated scenes, and designer-created scenes, including 1.96M 3D objects and covering 15 common scene types and 288 object classes.We particularly preserve massive small items in the scenes, resulting in realistic and complex layouts with an average of 41.5 objects per region.Our comprehensive data processing pipeline ensures simulatability by creating real-to-sim replicas for real-world scans, enhances interactivity by incorporating interactive objects into these scenes, and resolves object collisions by physical simulations.We demonstrate the value of InternScenes with two benchmark applications: scene layout generation and point-goal navigation. Both show the new challenges posed by the complex and realistic layouts. More importantly, InternScenes paves the way for scaling up the model training for both tasks, making the generation and navigation in such complex scenes possible. We commit to open-sourcing the data and benchmarks to benefit the whole community.

We introduce AnySplat, a feed-forward network for novel-view synthesis from uncalibrated image collections. In contrast to traditional neural-rendering pipelines that demand known camera poses and per-scene optimization, or recent feed-forward methods that buckle under the computational weight of dense views—our model predicts everything in one shot. A single forward pass yields a set of 3D Gaussian primitives encoding both scene geometry and appearance, and the corresponding camera intrinsics and extrinsics for each input image. This unified design scales effortlessly to casually captured, multi-view datasets without any pose annotations. In extensive zero-shot evaluations, AnySplat matches the quality of pose-aware baselines in both sparse- and dense-view scenarios while surpassing existing pose-free approaches. Moreover, it greatly reduces rendering latency compared to optimization-based neural fields, bringing real-time novel-view synthesis within reach for unconstrained capture settings. Project page: https://city-super.github.io/anysplat/.

We address the challenge of content diversity and controllability in pedestrian simulation for driving scenarios. Recent pedestrian animation frameworks have a significant limitation wherein they primarily focus on either following trajectory [48] or the content of the reference video [60], consequently overlooking the potential diversity of human motion within such scenarios. This limitation restricts the ability to generate pedestrian behaviors that exhibit a wider range of variations and realistic motions and therefore re-stricts its usage to provide rich motion content for other components in the driving simulation system, e.g., suddenly changed motion to which the autonomous vehicle should respond. In our approach, we strive to surpass the limitation by showcasing diverse human motions obtained from various sources, such as generated human motions, in ad-dition to following the given trajectory. The fundamental contribution of our framework lies in combining the motion tracking task with trajectory following, which enables the tracking of specific motion parts (e.g., upper body) while simultaneously following the given trajectory by a single policy. This way, we significantly enhance both the diver-sity of simulated human motion within the given scenario and the controllability of the content, including language-based control. Our framework facilitates the generation of a wide range of human motions, contributing to greater re-alism and adaptability in pedestrian simulations for driving scenarios.

Neural rendering methods have significantly advanced photo-realistic 3D scene rendering in various academic and industrial applications. The recent 3D Gaussian Splatting method has achieved the state-of-the-art rendering quality and speed combining the benefits of both primitive-based representations and volumetric representations. However, it often leads to heavily redundant Gaussians that try to fit every training view, neglecting the underlying scene ge-ometry. Consequently, the resulting model becomes less robust to significant view changes, texture-less area and lighting effects. We introduce Scaffold-GS, which uses an-chor points to distribute local 3D Gaussians, and predicts their attributes on-the-fly based on viewing direction and distance within the view frustum. Anchor growing and pruning strategies are developed based on the importance of neural Gaussians to reliably improve the scene cover-age. We show that our method effectively reduces redun-dant Gaussians while delivering high-quality rendering. We also demonstrates an enhanced capability to accommodate scenes with varying levels-of-detail and view-dependent ob-servations, without sacrificing the rendering speed. Project page: https://city-super.github.iolscaffold-gsl.

Generating multiview images from a single view facilitates the rapid generation of a 3D mesh conditioned on a single image. Recent methods [31] that introduce 3D global representation into diffusion models have shown the potential to generate consistent multiviews, but they have reduced generation speed and face challenges in maintaining generalizability and quality. To address this issue, we propose EpiDiff, a localized interactive multiview diffusion model. At the core of the proposed approach is to insert a lightweight epipolar attention block into the frozen diffusion model, leveraging epipolar constraints to enable cross-view interaction among feature maps of neighboring views. The newly initialized 3D modeling module preserves the original feature distribution of the diffusion model, exhibiting compatibility with a variety of base diffusion models. Experiments show that EpiDiff generates 16 multiview images in just 12 seconds, and it surpasses previous methods in quality evaluation metrics, including PSNR, SSIM and LPIPS. Additionally, EpiDiff can generate a more diverse distribution of views, improving the reconstruction quality from generated multiviews. Please see the project page at huanngzh.github.io/EpiDiff/.

In the evolving landscape of digital media and video pro-duction, the precise manipulation and reproduction of visual elements like camera movements and character actions are highly desired. Existing SLAM methods face limitations in dynamic scenes and human pose estimation often focuses on 2D projections, neglecting 3D statuses. To address these is-sues, we first introduce a reverse filming behavior estimation technique. It optimizes camera trajectories by leveraging NeRF as a differentiable renderer and refining SMPL tracks. We then introduce a cinematic transfer pipeline that is able to transfer various shot types to a new 2D video or a 3D vir-tual environment. The incorporation of 3D engine workflow enables superior rendering and control abilities, which also achieves a higher rating in the user study.

Creating and animating 3D biped cartoon characters is crucial and valuable in various applications. Compared with geometry, the diverse texture design plays an important role in making 3D biped cartoon characters vivid and charming. Therefore, we focus on automatic texture design for cartoon characters based on input instructions. This is challenging for domain-specific requirements and a lack of high-quality data. To address this challenge, we propose Make-It-Vivid, the first attempt to enable high-quality texture generation from text in UV space. We prepare a detailed text-texture paired data for 3D characters by using vision-question-answering agents. Then we customize a pretrained text-to-image model to generate texture map with template structure while preserving the natural 2D image knowledge. Furthermore, to enhance fine-grained details, we propose a novel adversarial learning scheme to shorten the domain gap between original dataset and realistic texture domain. Extensive experiments show that our approach outperforms current texture generation methods, resulting in efficient character texturing and faithful generation with prompts. Besides, we showcase various applications such as out of domain generation and texture stylization. We also provide an efficient generation system for automatic text-guided textured character generation and animation.

In this paper, we introduce the first large-scale video prediction model in the autonomous driving discipline. To eliminate the restriction of high-cost data collection and empower the generalization ability of our model, we ac-quire massive data from the web and pair it with diverse and high-quality text descriptions. The resultant dataset accumulates over 2000 hours of driving videos, spanning areas all over the world with diverse weather conditions and traffic scenarios. Inheriting the merits from recent latent diffusion models, our model, dubbed GenAD, handles the challenging dynamics in driving scenes with novel tem-poral reasoning blocks. We showcase that it can general-ize to various unseen driving datasets in a zero-shot man-ner, surpassing general or driving-specific video prediction counterparts. Furthermore, GenAD can be adapted into an action-conditioned prediction model or a motion planner, holding great potential for real-world driving applications.

Generating large-scale 3D scenes cannot simply apply existing 3D object synthesis technique since 3D scenes usually hold complex spatial configurations and consist of a number of objects at varying scales. We thus propose a practical and efficient 3D representation that incorporates an equivariant radiance field with the guidance of a bird's-eye view (BEV) map. Concretely, objects of synthesized 3D scenes could be easily manipulated through steering the corresponding BEV maps. Moreover, by adequately incorporating positional encoding and low-pass filters into the generator, the representation becomes equivariant to the given BEV map. Such equivariance allows us to produce large-scale, even infinite-scale, 3D scenes via synthesizing local scenes and then stitching them with smooth consistency. Extensive experiments on 3D scene datasets demonstrate the effectiveness of our approach. Our project web-site is at: https://zqh0253.github.io/BerfScene/.

Diffusion models have achieved remarkable image generation quality surpassing previous generative models. However, a notable limitation of diffusion models, in comparison to GANs, is their difficulty in smoothly interpolating between two image samples, due to their highly unstructured latent space. Such a smooth interpolation is intriguing as it naturally serves as a solution for the image mor-phing task with many applications. In this work, we address this limitation via DiffMorpher, an approach that enables smooth and natural image interpolation by harnessing the prior knowledge of a pretrained diffusion model. Our key idea is to capture the semantics of the two images by fitting two LoRAs to them respectively, and interpolate between both the LoRA parameters and the latent noises to ensure a smooth semantic transition, where correspon-dence automatically emerges without the need for annotation. In addition, we propose an attention interpolation and injection technique, an adaptive normalization adjustment method, and a new sampling schedule to further enhance the smoothness between consecutive images. Extensive experiments demonstrate that DiffMorpher achieves starkly better image morphing effects than previous methods across a variety of object categories, bridging a critical functional gap that distinguished diffusion models from GANs.

Pre-training a model and then fine-tuning it on down-stream tasks has demonstrated significant success in the 2D image and NLP domains. However, due to the unordered and non-uniform density characteristics of point clouds, it is non-trivial to explore the prior knowledge of point clouds and pre-train a point cloud backbone. In this paper, we propose a novel pre-training method called Point cloud Diffusion pre-training (PointDif). We consider the point cloud pre-training task as a conditional point-to-point gen-eration problem and introduce a conditional point genera-tor. This generator aggregates the features extracted by the backbone and employs them as the condition to guide the point-to-point recovery from the noisy point cloud, thereby assisting the backbone in capturing both local and global geometric priors as well as the global point density distri-bution of the object. We also present a recurrent uniform sampling optimization strategy, which enables the model to uniformly recover from various noise levels and learn from balanced supervision. Our PointDif achieves substan-tial improvement across various real-world datasets for di-verse downstream tasks such as classification, segmentation and detection. Specifically, PointDif attains 70.0% mIoU on S3DIS Area 5 for the segmentation task and achieves an average improvement of 2.4% on ScanObjectNN for the classification task compared to TAP. Furthermore, our pre-training framework can be flexibly applied to diverse point cloud backbones and bring considerable gains. Code is available at https://github.com/zhengxiaozx/PointDif.

With the advance of text-to-image (T2I) diffusion models (e.g., Stable Diffusion) and corresponding personalization techniques such as DreamBooth and LoRA, everyone can manifest their imagination into high-quality images at an affordable cost. However, adding motion dynamics to existing high-quality personalized T2Is and enabling them to generate animations remains an open challenge. In this paper, we present AnimateDiff, a practical framework for animating personalized T2I models without requiring model-specific tuning. At the core of our framework is a plug-and-play motion module that can be trained once and seamlessly integrated into any personalized T2Is originating from the same base T2I. Through our proposed training strategy, the motion module effectively learns transferable motion priors from real-world videos. Once trained, the motion module can be inserted into a personalized T2I model to form a personalized animation generator. We further propose MotionLoRA, a lightweight fine-tuning technique for AnimateDiff that enables a pre-trained motion module to adapt to new motion patterns, such as different shot types, at a low training and data collection cost. We evaluate AnimateDiff and MotionLoRA on several public representative personalized T2I models collected from the community. The results demonstrate that our approaches help these models generate temporally smooth animation clips while preserving the visual quality and motion diversity. Codes and pre-trained weights are available at https://github.com/guoyww/AnimateDiff.

Human-Scene Interaction (HSI) is a vital component of fields like embodied AI and virtual reality. Despite advancements in motion quality and physical plausibility, two pivotal factors, versatile interaction control and the development of a user-friendly interface, require further exploration before the practical application of HSI. This paper presents a unified HSI framework, UniHSI, which supports unified control of diverse interactions through language commands. The framework defines interaction as ``Chain of Contacts (CoC)", representing steps involving human joint-object part pairs. This concept is inspired by the strong correlation between interaction types and corresponding contact regions. Based on the definition, UniHSI constitutes a Large Language Model (LLM) Planner to translate language prompts into task plans in the form of CoC, and a Unified Controller that turns CoC into uniform task execution. To facilitate training and evaluation, we collect a new dataset named ScenePlan that encompasses thousands of task plans generated by LLMs based on diverse scenarios. Comprehensive experiments demonstrate the effectiveness of our framework in versatile task execution and generalizability to real scanned scenes.

Text-conditioned motion synthesis has made remarkable progress with the emergence of diffusion models. However, the majority of these motion diffusion models are primarily designed for a single character and overlook multi-human interactions. In our approach, we strive to explore this problem by synthesizing human motion with interactions for a group of characters of any size in a zero-shot manner. The key aspect of our approach is the adaptation of human-wise interactions as pairs of human joints that can be either in contact or separated by a desired distance. In contrast to existing methods that necessitate training motion generation models on multi-human motion datasets with a fixed number of characters, our approach inherently possesses the flexibility to model human interactions involving an arbitrary number of individuals, thereby transcending the limitations imposed by the training data. We introduce a novel controllable motion generation method, InterControl, to encourage the synthesized motions maintaining the desired distance between joint pairs. It consists of a motion controller and an inverse kinematics guidance module that realistically and accurately aligns the joints of synthesized characters to the desired location. Furthermore, we demonstrate that the distance between joint pairs for human-wise interactions can be generated using an off-the-shelf Large Language Model (LLM). Experimental results highlight the capability of our framework to generate interactions with multiple human characters and its potential to work with off-the-shelf physics-based character simulators. Code is available at https://github.com/zhenzhiwang/intercontrol.

Human image animation involves generating videos from a character photo, allowing user control and unlocking the potential for video and movie production. While recent approaches yield impressive results using high-quality training data, the inaccessibility of these datasets hampers fair and transparent benchmarking. Moreover, these approaches prioritize 2D human motion and overlook the significance of camera motions in videos, leading to limited control and unstable video generation. To demystify the training data, we present HumanVid, the first large-scale high-quality dataset tailored for human image animation, which combines crafted real-world and synthetic data. For the real-world data, we compile a vast collection of real-world videos from the internet. We developed and applied careful filtering rules to ensure video quality, resulting in a curated collection of 20K high-resolution (1080P) human-centric videos. Human and camera motion annotation is accomplished using a 2D pose estimator and a SLAM-based method. To expand our synthetic dataset, we collected 10K 3D avatar assets and leveraged existing assets of body shapes, skin textures and clothings. Notably, we introduce a rule-based camera trajectory generation method, enabling the synthetic pipeline to incorporate diverse and precise camera motion annotation, which can rarely be found in real-world data. To verify the effectiveness of HumanVid, we establish a baseline model named CamAnimate, short for Camera-controllable Human Animation, that considers both human and camera motions as conditions. Through extensive experimentation, we demonstrate that such simple baseline training on our HumanVid achieves state-of-the-art performance in controlling both human pose and camera motions, setting a new benchmark. Demo, data and code could be found in the project website: https://humanvid.github.io/.

Recent advancements in 3D generation have leveraged synthetic datasets with ground truth 3D assets and predefined camera trajectories. However, the potential of adopting real-world datasets, which can produce significantly more realistic 3D scenes, remains largely unexplored. In this work, we delve into the key challenge of the complex and scene-specific camera trajectories found in real-world captures. We introduce Director3D, a robust open-world text-to-3D generation framework, designed to generate both real-world 3D scenes and adaptive camera trajectories. To achieve this, (1) we first utilize a Trajectory Diffusion Transformer, acting as the \emph{Cinematographer}, to model the distribution of camera trajectories based on textual descriptions. Next, a Gaussian-driven Multi-view Latent Diffusion Model serves as the \emph{Decorator}, modeling the image sequence distribution given the camera trajectories and texts. This model, fine-tuned from a 2D diffusion model, directly generates pixel-aligned 3D Gaussians as an immediate 3D scene representation for consistent denoising. Lastly, the 3D Gaussians are further refined by a novel SDS++ loss as the \emph{Detailer}, which incorporates the prior of the 2D diffusion model. Extensive experiments demonstrate that Director3D outperforms existing methods, offering superior performance in real-world 3D generation.

Garment animation is ubiquitous in various applications, such as virtual reality, gaming, and film producing. Recently, learning-based approaches obtain compelling performance in animating diverse garments under versatile scenarios. Nevertheless, to mimic the deformations of the observed garments, data-driven methods require large scale of garment data, which are both resource-wise expensive and time-consuming. In addition, forcing models to match the dynamics of observed garment animation may hinder the potentials to generalize to unseen cases. In this paper, instead of using garment-wise supervised-learning we adopt a disentangled scheme to learn how to animate observed garments: 1). learning constitutive behaviors from the observed cloth; 2). dynamically animate various garments constrained by the learned constitutive laws. Specifically, we propose Energy Unit network (EUNet) to model the constitutive relations in the format of energy. Without the priors from analytical physics models and differentiable simulation engines, EUNet is able to directly capture the constitutive behaviors from the observed piece of cloth and uniformly describes the change of energy caused by deformations, such as stretching and bending. We further apply the pre-trained EUNet to animate various garments based on energy optimizations. The disentangled scheme alleviates the need of garment data and enables us to utilize the dynamics of a piece of cloth for animating garments. Experiments show that while EUNet effectively delivers the energy gradients due to the deformations, models constrained by EUNet achieve more stable and physically plausible performance comparing with those trained in garment-wise supervised manner.

Representing 3D scenes from multiview images remains a core challenge in computer vision and graphics, requiring both reliable rendering and reconstruction, which often conflicts due to the mismatched prioritization of image quality over precise underlying scene geometry. Although both neural implicit surfaces and explicit Gaussian primitives have advanced with neural rendering techniques, current methods impose strict constraints on density fields or primitive shapes, which enhances the affinity for geometric reconstruction at the sacrifice of rendering quality. To address this dilemma, we introduce GSDF, a dual-branch architecture combining 3D Gaussian Splatting (3DGS) and neural Signed Distance Fields (SDF). Our approach leverages mutual guidance and joint supervision during the training process to mutually enhance reconstruction and rendering. Specifically, our method guides the Gaussian primitives to locate near potential surfaces and accelerates the SDF convergence. This implicit mutual guidance ensures robustness and accuracy in both synthetic and real-world scenarios. Experimental results demonstrate that our method boosts the SDF optimization process to reconstruct more detailed geometry, while reducing floaters and blurry edge artifacts in rendering by aligning Gaussian primitives with the underlying geometry.

Boosting Normal BoostingFig. 1.Overview.Given a single image, the existing image-to-3D generative models always synthesize 3D meshes with flawed geometry and RGB textures only.Our method not only boosts existing approaches with PBR materials, empowering relighting under various lighting conditions, but also boosts the object's normal maps, capturing more intricate details and better aligning with the given image.Notably, we fine-tune Stable Diffusion to estimate the albedo map from the single-view RGB image and lift it to multi-view albedo maps for a complete albedo UV.

Digital human motion synthesis is a vibrant research field with applications in movies, AR/VR, and video games. Whereas methods were proposed to generate natural and realistic human motions, most only focus on modeling humans and largely ignore object movements. Generating task-oriented human-object interaction motions in simulation is challenging. For different intents of using the objects, humans conduct various motions, which requires the human first to approach the objects and then make them move consistently with the human instead of staying still. Also, to deploy in downstream applications, the synthesized motions are desired to be flexible in length, providing options to personalize the predicted motions for various purposes. To this end, we propose TOHO: Task-Oriented Human-Object Interactions Generation with Implicit Neural Representations, which generates full human-object interaction motions to conduct specific tasks, given only the task type, the object, and a starting human status. TOHO generates human-object motions in four steps: 1) it first estimates the object’s final position given the task intent; 2) it then generates keyframe poses grasping the objects; 3) after that, it infills the keyframes and generates continuous motions; 4) finally, it applies a compact closed-form object motion estimation to generate the object motion. Our method generates continuous motions that are parameterized only by the temporal coordinate, which allows for upsampling of the sequence to arbitrary frames and adjusting the motion speeds by designing the temporal coordinate vector. This work takes a step further toward general human-scene interaction simulation.

Portrait stylization is a long-standing task enabling extensive applications. Although 2D-based methods have made great progress in recent years, real-world applications such as metaverse and games often demand 3D content. On the other hand, the requirement of 3D data, which is costly to acquire, significantly impedes the development of 3D portrait stylization methods. In this paper, inspired by the success of 3D-aware GANs that bridge 2D and 3D domains with 3D fields as the intermediate representation for rendering 2D images, we propose a novel method, dubbed HyperStyle3D, based on 3D-aware GANs for 3D portrait stylization. At the core of our method is a hyper-network learned to manipulate the parameters of the generator in a single forward pass. It not only offers a strong capacity to handle multiple styles with a single model, but also enables flexible fine-grained stylization that affects only texture, shape, or local part of the portrait. While the use of 3D-aware GANs bypasses the requirement of 3D data, we further alleviate the necessity of style images with the CLIP model being the style guidance. We conduct an extensive set of experiments across the style, attribute, and shape, and meanwhile, measure the 3D consistency. These experiments demonstrate the superior capability of our HyperStyle3D model in rendering 3D-consistent images in diverse styles, deforming the face shape, and editing various attributes.

Real-time character control is an essential component for interactive experiences, with a broad range of applications, including physics simulations, video games, and virtual reality. The success of diffusion models for image synthesis has led to the use of these models for motion synthesis. However, the majority of these motion diffusion models are primarily designed for offline applications, where space-time models are used to synthesize an entire sequence of frames simultaneously with a pre-specified length. To enable real-time motion synthesis with diffusion model that allows time-varying controls, we propose A-MDM (Auto-regressive Motion Diffusion Model). Our conditional diffusion model takes an initial pose as input, and auto-regressively generates successive motion frames conditioned on the previous frame. Despite its streamlined network architecture, which uses simple MLPs, our framework is capable of generating diverse, long-horizon, and high-fidelity motion sequences. Furthermore, we introduce a suite of techniques for incorporating interactive controls into A-MDM, such as task-oriented sampling, in-painting, and hierarchical reinforcement learning (See Figure 1). These techniques enable a pre-trained A-MDM to be efficiently adapted for a variety of new downstream tasks. We conduct a comprehensive suite of experiments to demonstrate the effectiveness of A-MDM, and compare its performance against state-of-the-art auto-regressive methods.

Existing image restoration methods mostly leverage the posterior distribution of natural images. However, they often assume known degradation and also require supervised training, which restricts their adaptation to complex real applications. In this work, we propose the Generative Diffusion Prior (GDP) to effectively model the posterior distributions in an unsupervised sampling manner. GDP utilizes a pre-train denoising diffusion generative model (DDPM) for solving linear inverse, non-linear, or blind problems. Specifically, GDP systematically explores a protocol of conditional guidance, which is verified more practical than the commonly used guidance way. Furthermore, GDP is strength at optimizing the parameters of degradation model during the denoising process, achieving blind image restoration. Besides, we devise hierarchical guidance and patch-based methods, enabling the GDP to generate images of arbitrary resolutions. Experimentally, we demonstrate GDP's versatility on several image datasets for linear problems, such as super-resolution, deblurring, inpainting, and colorization, as well as non-linear and blind issues, such as low-light enhancement and HDR image recovery. GDP outperforms the current leading unsupervised methods on the diverse benchmarks in reconstruction quality and perceptual quality. Moreover, GDP also generalizes well for natural images or synthesized images with arbitrary sizes from various tasks out of the distribution of the ImageNet training set. The project page is available at https://generativediffusionprior.github.io/

StyleGAN has achieved great progress in 2D face reconstruction and semantic editing via image inversion and latent editing. While studies over extending 2D StyleGAN to 3D faces have emerged, a corresponding generic 3D GAN inversion framework is still missing, limiting the applications of 3D face reconstruction and semantic editing. In this paper, we study the challenging problem of 3D GAN inversion where a latent code is predicted given a single face image to faithfully recover its 3D shapes and detailed textures. The problem is ill-posed: innumerable compositions of shape and texture could be rendered to the current image. Furthermore, with the limited capacity of a global latent code, 2D inversion methods cannot preserve faithful shape and texture at the same time when applied to 3D models. To solve this problem, we devise an effective self-training scheme to constrain the learning of inversion. The learning is done efficiently without any real-world 2D-3D training pairs but proxy samples generated from a 3D GAN. In addition, apart from a global latent code that captures the coarse shape and texture information, we augment the generation network with a local branch, where pixel-aligned features are added to faithfully reconstruct face details. We further consider a new pipeline to perform 3D view-consistent editing. Extensive experiments show that our method outperforms state-of-the-art inversion methods in both shape and texture reconstruction quality.

Mesh generation is of great value in various applications involving computer graphics and virtual content, yet designing generative models for meshes is challenging due to their irregular data structure and inconsistent topology of meshes in the same category. In this work, we design a novel sparse latent point diffusion model for mesh generation. Our key insight is to regard point clouds as an intermediate representation of meshes, and model the distribution of point clouds instead. While meshes can be generated from point clouds via techniques like Shape as Points (SAP), the challenges of directly generating meshes can be effectively avoided. To boost the efficiency and controllability of our mesh generation method, we propose to further encode point clouds to a set of sparse latent points with pointwise semantic meaningful features, where two DDPMs are trained in the space of sparse latent points to respectively model the distribution of the latent point positions and features at these latent points. We find that sampling in this latent space is faster than directly sampling dense point clouds. Moreover, the sparse latent points also enable us to explicitly control both the overall structures and local details of the generated meshes. Extensive experiments are conducted on the ShapeNet dataset, where our proposed sparse latent point diffusion model achieves superior performance in terms of generation quality and controllability when compared to existing methods. Project page, code and appendix: https://slide-3d.github.io.

Purely MLP-based neural radiance fields (NeRF-based methods) often suffer from underfitting with blurred renderings on large-scale scenes due to limited model capacity. Recent approaches propose to geographically divide the scene and adopt multiple sub-NeRFs to model each region individually, leading to linear scale-up in training costs and the number of sub-NeRFs as the scene expands. An alternative solution is to use a feature grid representation, which is computationally efficient and can naturally scale to a large scene with increased grid resolutions. However, the feature grid tends to be less constrained and often reaches suboptimal solutions, producing noisy artifacts in renderings, especially in regions with complex geometry and texture. In this work, we present a new framework that realizes high-fidelity rendering on large urban scenes while being computationally efficient. We propose to use a compact multi-resolution ground feature plane representation to coarsely capture the scene, and complement it with positional encoding inputs through another NeRF branch for rendering in a joint learning fashion. We show that such an integration can utilize the advantages of two alternative solutions: a light-weighted NeRF is sufficient, under the guidance of the feature grid representation, to render photorealistic novel views with fine details; and the jointly optimized ground feature planes, can meanwhile gain further refinements, forming a more accurate and compact feature space and output much more natural rendering results.

Neural radiance fields (NeRF) and its subsequent variants have led to remarkable progress in neural rendering. While most of recent neural rendering works focus on objects and small-scale scenes, developing neural rendering methods for city-scale scenes is of great potential in many real-world applications. However, this line of research is impeded by the absence of a comprehensive and high-quality dataset, yet collecting such a dataset over real city-scale scenes is costly, sensitive, and technically infeasible. To this end, we build a large-scale, comprehensive, and high-quality synthetic dataset for city-scale neural rendering researches. Leveraging the Unreal Engine 5 City Sample project, we developed a pipeline to easily collect aerial and street city views, accompanied by ground-truth camera poses and a range of additional data modalities. Flexible controls on environmental factors like light, weather, human and car crowd are also available in our pipeline, supporting the need of various tasks covering city-scale neural rendering and beyond. The resulting pilot dataset, MatrixCity, contains 67k aerial images and 452k street images from two city maps of total size 28km2. On top of MatrixCity, a thorough benchmark is also conducted, which not only reveals unique challenges of the task of city-scale neural rendering, but also highlights potential improvements for future works. The dataset and code will be publicly available at the project page: https://city-super.github.io/matrixcity/.

Realistic human-centric rendering plays a key role in both computer vision and computer graphics. Rapid progress has been made in the algorithm aspect over the years, yet existing human-centric rendering datasets and benchmarks are rather impoverished in terms of diversity (e.g., outfit's fabric/material, body's interaction with objects, and motion sequences), which are crucial for rendering effect. Researchers are usually constrained to explore and evaluate a small set of rendering problems on current datasets, while real-world applications require methods to be robust across different scenarios. In this work, we present DNA-Rendering, a large-scale, high-fidelity repository of human performance data for neural actor rendering. DNA-Rendering presents several appealing attributes. First, our dataset contains over 1500 human subjects, 5000 motion sequences, and 67.5M frames' data volume. Upon the massive collections, we provide human subjects with grand categories of pose actions, body shapes, clothing, accessories, hairdos, and object intersection, which ranges the geometry and appearance variances from everyday life to professional occasions. Second, we provide rich assets for each subject – 2D/3D human body keypoints, foreground masks, SMPLX models, cloth/accessory materials, multi-view images, and videos. These assets boost the current method's accuracy on downstream rendering tasks. Third, we construct a professional multi-view system to capture data, which contains 60 synchronous cameras with max 4096 × 3000 resolution, 15 fps speed, and stern camera calibration steps, ensuring high-quality resources for task training and evaluation.Along with the dataset, we provide a large-scale and quantitative benchmark in full-scale, with multiple tasks to evaluate the existing progress of novel view synthesis, novel pose animation synthesis, and novel identity rendering methods. In this manuscript, we describe our DNA-Rendering effort as a revealing of new observations, challenges, and future directions to human-centric rendering. The dataset, code, and benchmarks will be publicly available at https://dna-rendering.github.io/.

We present a new method for generating realistic and view-consistent images with fine geometry from 2D image collections. Our method proposes a hybrid explicit-implicit representation called OrthoPlanes, which encodes fine-grained 3D information in feature maps that can be efficiently generated by modifying 2D StyleGANs. Compared to previous representations, our method has better scalability and expressiveness with clear and explicit information. As a result, our method can handle more challenging view-angles and synthesize articulated objects with high spatial degree of freedom. Experiments demonstrate that our method achieves state-of-the-art results on FFHQ and SHHQ datasets, both quantitatively and qualitatively. Project page: https://orthoplanes.github.io/.

Mimicking realistic dynamics in 3D garment animations is a challenging task due to the complex nature of multi-layered garments and the variety of outer forces involved. Existing approaches mostly focus on single-layered garments driven by only human bodies and struggle to handle general scenarios. In this paper, we propose a novel data-driven method, called LayersNet, to model garment-level animations as particle-wise interactions in a micro physics system. We improve simulation efficiency by representing garments as patch-level particles in a two-level structural hierarchy. Moreover, we introduce a novel Rotation Equivalent Transformation with Rotation Invariant Attention that leverage the rotation invariance and additivity of physics systems to better model outer forces. To verify the effectiveness of our approach and bridge the gap between experimental environments and real-world scenarios, we introduce a new challenging dataset, D-LAYERS, containing 700K frames of dynamics of 4,900 combinations of multi-layered garments driven by human bodies and randomly sampled wind. Our LayersNet achieves superior performance both quantitatively and qualitatively. Project page: www.mmlab-ntu.com/project/layersnet/index.html.

Both indoor and outdoor environments are inherently structured and repetitive. Traditional modeling pipelines keep an asset library storing unique object templates, which is both versatile and memory efficient in practice. Inspired by this observation, we propose AssetField, a novel neural scene representation that learns a set of object-aware ground feature planes to represent the scene, where an asset library storing template feature patches can be constructed in an unsupervised manner. Unlike existing methods which require object masks to query spatial points for object editing, our ground feature plane representation offers a natural visualization of the scene in the bird-eye view, allowing a variety of operations (e.g. translation, duplication, deformation) on objects to configure a new scene. With the template feature patches, group editing is enabled for scenes with many recurring items to avoid repetitive work on object individuals. We show that AssetField not only achieves competitive performance for novel-view synthesis but also generates realistic renderings for new scene configurations.

Synthetic data has emerged as a promising source for 3D human research as it offers low-cost access to large-scale human datasets. To advance the diversity and annotation quality of human models, we introduce a new synthetic dataset, SynBody, with three appealing features: 1) a clothed parametric human model that can generate a diverse range of subjects; 2) the layered human representation that naturally offers high-quality 3D annotations to support multiple tasks; 3) a scalable system for producing realistic data to facilitate real-world tasks. The dataset comprises 1.2M images with corresponding accurate 3D annotations, covering 10,000 human body models, 1,187 actions, and various viewpoints. The dataset includes two subsets for human pose and shape estimation as well as human neural rendering. Extensive experiments on SynBody indicate that it substantially enhances both SMPL and SMPL-X estimation. Furthermore, the incorporation of layered annotations offers a valuable training resource for investigating the Human Neural Radiance Fields(NeRF).

We present 3DHumanGAN, a 3D-aware generative adversarial network that synthesizes photo-like images of full-body humans with consistent appearances under different view-angles and body-poses. To tackle the representational and computational challenges in synthesizing the articulated structure of human bodies, we propose a novel generator architecture in which a 2D convolutional backbone is modulated by a 3D pose mapping network. The 3D pose mapping network is formulated as a renderable implicit function conditioned on a posed 3D human mesh. This design has several merits: i) it leverages the strength of 2D GANs to produce high-quality images; ii) it generates consistent images under varying view-angles and poses; iii) the model can incorporate the 3D human prior and enable pose conditioning. Our model is adversarially learned from a collection of web images needless of manual annotation.

This work presents an easy-to-use regularizer for GAN training, which helps explicitly link some axes of the latent space to a set of pixels in the synthesized image. Establishing such a connection facilitates a more convenient local control of GAN generation, where users can alter the image content only within a spatial area simply by partially resampling the latent code. Experimental results confirm four appealing properties of our regularizer, which we call LinkGAN. (1) The latent-pixel linkage is applicable to either a fixed region (i.e., same for all instances) or a particular semantic category (i.e., varying across instances), like the sky. (2) Two or multiple regions can be independently linked to different latent axes, which further supports joint control. (3) Our regularizer can improve the spatial controllability of both 2D and 3D-aware GAN models, barely sacrificing the synthesis performance. (4) The models trained with our regularizer are compatible with GAN inversion techniques and maintain editability on real images. Project page can be found here.

Factor model is a fundamental investment tool in quantitative investment, which can be empowered by deep learning to become more flexible and efficient in practical complicated investing situations. However, it is still an open question to build a factor model that can conduct stock prediction in an online and adaptive setting, where the model can adapt itself to match the current market regime identified based on only point-in-time market information. To tackle this problem, we propose the first deep learning based online and adaptive factor model, HireVAE, at the core of which is a hierarchical latent space that embeds the underlying relationship between the market situation and stock-wise latent factors, so that HireVAE can effectively estimate useful latent factors given only historical market information and subsequently predict accurate stock returns. Across four commonly used real stock market benchmarks, the proposed HireVAE demonstrate superior performance in terms of active returns over previous methods, verifying the potential of such online and adaptive factor model.

Synthesizing high-fidelity head avatars is a central problem for computer vision and graphics. While head avatar synthesis algorithms have advanced rapidly, the best ones still face great obstacles in real-world scenarios. One of the vital causes is the inadequate datasets -- 1) current public datasets can only support researchers to explore high-fidelity head avatars in one or two task directions; 2) these datasets usually contain digital head assets with limited data volume, and narrow distribution over different attributes, such as expressions, ages, and accessories. In this paper, we present RenderMe-360, a comprehensive 4D human head dataset to drive advance in head avatar algorithms across different scenarios. It contains massive data assets, with 243+ million complete head frames and over 800k video sequences from 500 different identities captured by multi-view cameras at 30 FPS. It is a large-scale digital library for head avatars with three key attributes: 1) High Fidelity: all subjects are captured in 360 degrees via 60 synchronized, high-resolution 2K cameras. 2) High Diversity: The collected subjects vary from different ages, eras, ethnicities, and cultures, providing abundant materials with distinctive styles in appearance and geometry. Moreover, each subject is asked to perform various dynamic motions, such as expressions and head rotations, which further extend the richness of assets. 3) Rich Annotations: the dataset provides annotations with different granularities: cameras' parameters, background matting, scan, 2D/3D facial landmarks, FLAME fitting, and text description. Based on the dataset, we build a comprehensive benchmark for head avatar research, with 16 state-of-the-art methods performed on five main tasks: novel view synthesis, novel expression synthesis, hair rendering, hair editing, and talking head generation. Our experiments uncover the strengths and flaws of state-of-the-art methods. RenderMe-360 opens the door for future exploration in modern head avatars. All of the data, code, and models will be publicly available at https://renderme-360.github.io/.

A good metric, which promises a reliable comparison between solutions, is essential for any well-defined task. Unlike most vision tasks that have per-sample ground-truth, image synthesis tasks target generating unseen data and hence are usually evaluated through a distributional distance between one set of real samples and another set of generated samples. This study presents an empirical investigation into the evaluation of synthesis performance, with generative adversarial networks (GANs) as a representative of generative models. In particular, we make in-depth analyses of various factors, including how to represent a data point in the representation space, how to calculate a fair distance using selected samples, and how many instances to use from each set. Extensive experiments conducted on multiple datasets and settings reveal several important findings. Firstly, a group of models that include both CNN-based and ViT-based architectures serve as reliable and robust feature extractors for measurement evaluation. Secondly, Centered Kernel Alignment (CKA) provides a better comparison across various extractors and hierarchical layers in one model. Finally, CKA is more sample-efficient and enjoys better agreement with human judgment in characterizing the similarity between two internal data correlations. These findings contribute to the development of a new measurement system, which enables a consistent and reliable re-evaluation of current state-of-the-art generative models.

The success of style-based generators largely benefits from style modulation, which helps take care of the cross-instance variation within data. However, the instance-wise stochasticity is typically introduced via regular convolution, where kernels interact with features at some fixed locations, limiting its capacity for modeling geometric variation. To alleviate this problem, we equip the generator in generative adversarial networks (GANs) with a plug-and-play module, termed as modulated transformation module (MTM). This module predicts spatial offsets under the control of latent codes, based on which the convolution operation can be applied at variable locations for different instances, and hence offers the model an additional degree of freedom to handle geometry deformation. Extensive experiments suggest that our approach can be faithfully generalized to various generative tasks, including image generation, 3D-aware image synthesis, and video generation, and get compatible with state-of-the-art frameworks without any hyper-parameter tuning. It is noteworthy that, towards human generation on the challenging TaiChi dataset, we improve the FID of StyleGAN3 from 21.36 to 13.60, demonstrating the efficacy of learning modulated geometry transformation. Code and models are available at https://github.com/limbo0000/mtm.

The ability to synthesize long-term human motion sequences in real-world scenes can facilitate numerous applications. Previous approaches for scene-aware motion synthesis are constrained by pre-defined target objects or positions and thus limit the diversity of human-scene interactions for synthesized motions. In this paper, we focus on the problem of synthesizing diverse scene-aware human motions under the guidance of target action sequences. To achieve this, we first decompose the diversity of scene-aware human motions into three aspects, namely interaction diversity (e.g. sitting on different objects with different poses in the given scenes), path diversity (e.g. moving to the target locations following different paths), and the motion diversity (e.g. having various body movements during moving). Based on this factorized scheme, a hierarchical framework is proposed, with each sub-module responsible for modeling one aspect. We assess the effectiveness of our framework on two challenging datasets for scene-aware human motion synthesis. The experiment results show that the proposed framework remarkably outperforms previous methods in terms of diversity and naturalness.

Human skeleton, as a compact representation of human action, has received increasing attention in recent years. Many skeleton-based action recognition methods adopt GCNs to extract features on top of human skeletons. Despite the positive results shown in these attempts, GCN-based methods are subject to limitations in robustness, interoperability, and scalability. In this work, we propose PoseConv3D, a new approach to skeleton-based action recognition. PoseConv3D relies on a 3D heatmap volume instead of a graph sequence as the base representation of human skeletons. Compared to GCN-based methods, PoseConv3D is more effective in learning spatiotemporal features, more robust against pose estimation noises, and generalizes better in cross-dataset settings. Also, PoseConv3D can handle multiple-person scenarios without additional computation costs. The hierarchical features can be easily integrated with other modalities at early fusion stages, providing a great design space to boost the performance. PoseConv3D achieves the state-of-the-art on five of six standard skeleton-based action recognition benchmarks. Once fused with other modalities, it achieves the state-of-the-art on all eight multi-modality action recognition benchmarks. Code has been made available at: https://github.com/kennymckormick/pyskl.

Generating speech-consistent body and gesture movements is a long-standing problem in virtual avatar creation. Previous studies often synthesize pose movement in a holistic manner, where poses of all joints are generated simultaneously. Such a straightforward pipeline fails to generate fine-grained co-speech gestures. One observation is that the hierarchical semantics in speech and the hierarchical structures of human gestures can be naturally described into multiple granularities and associated together. To fully utilize the rich connections between speech audio and human gestures, we propose a novel framework named Hierarchical Audio-to-Gesture (HA2G) for co-speech gesture generation. In HA2G, a Hierarchical Audio Learner extracts audio representations across semantic granularities. A Hierarchical Pose Inferer subsequently renders the entire human pose gradually in a hierarchical manner. To enhance the quality of synthesized gestures, we develop a contrastive learning strategy based on audio-text alignment for better audio representations. Extensive experiments and human evaluation demonstrate that the proposed method renders realistic co-speech gestures and out-performs previous methods in a clear margin. Project page: https://alvinliu0.github.io/projects/HA2G.

Semi-supervised action recognition is a challenging but important task due to the high cost of data annotation. A common approach to this problem is to assign unlabeled data with pseudo-labels, which are then used as additional supervision in training. Typically in recent work, the pseudo-labels are obtained by training a model on the labeled data, and then using confident predictions from the model to teach itself. In this work, we propose a more effective pseudo-labeling scheme, called Cross-Model Pseudo-Labeling (CMPL). Concretely, we introduce a lightweight auxiliary network in addition to the primary backbone, and ask them to predict pseudo-labels for each other. We observe that, due to their different structural biases, these two models tend to learn complementary representations from the same video clips. Each model can thus benefit from its counterpart by utilizing cross-model predictions as supervision. Experiments on different data partition protocols demonstrate the significant improvement of our framework over existing alternatives. For example, CMPL achieves 17.6% and 25.1% Top-1 accuracy on Kinetics-400 and UCF-101 using only the RGB modality and 1% labeled data, outperforming our baseline model, FixMatch [17], by 9.0% and 10.3%, respectively.11Project page is at https://justimyhxu.github.io/projects/cmpl/.

Recent advances like StyleGAN have promoted the growth of controllable facial editing. To address its core challenge of attribute decoupling in a single latent space, attempts have been made to adopt dual-space GAN for better disentanglement of style and content representations. Nonetheless, these methods are still incompetent to obtain plausible editing results with high controllability, especially for complicated attributes. In this study, we highlight the importance of interaction in a dual-space GAN for more controllable editing. We propose TransEditor, a novel Transformer-based framework to enhance such interaction. Besides, we develop a new dual-space editing and inversion strategy to provide additional editing flexibility. Extensive experiments demonstrate the superiority of the proposed framework in image quality and editing capability, suggesting the effectiveness of TransEditor for highly controllable facial editing. Code and models are publicly available at https://github.com/BillyXYB/TransEditor.

Discriminator plays a vital role in training generative adversarial networks (GANs) via distinguishing real and synthesized samples. While the real data distribution remains the same, the synthesis distribution keeps varying because of the evolving generator, and thus effects a corresponding change of the bi-classification task assigned to the discriminator. We argue that a discriminator with an on-the-fly adjustment on its capacity can better accommodate such a time-varying task. A comprehensive empirical study confirms that the proposed training strategy, termed as DynamicD, improves the synthesis performance without incurring any additional computation cost or training objectives. Two capacity adjusting schemes are developed for training GANs under different data regimes: i) given a sufficient amount of training data, the discriminator benefits from a progressively increased learning capacity, and ii) when the training data is limited, gradually decreasing the layer width mitigates the over-fitting issue of the discriminator. Experiments on both 2D and 3D-aware image synthesis tasks conducted on a range of datasets substantiate the generalizability of our DynamicD as well as its substantial improvement over the baselines. Furthermore, DynamicD is synergistic to other discriminator-improving approaches (including data augmentation, regularizers, and pre-training), and brings continuous performance gain when combined with them for learning GANs. Code will be made publicly available.

Learning a good image prior is a long-term goal for image restoration and manipulation. While existing methods like deep image prior (DIP) capture low-level image statistics, there are still gaps toward an image prior that captures rich image semantics including color, spatial coherence, textures, and high-level concepts. This work presents an effective way to exploit the image prior captured by a generative adversarial network (GAN) trained on large-scale natural images. As shown in Fig. 1, the deep generative prior (DGP) provides compelling results to restore missing semantics, e.g., color, patch, resolution, of various degraded images. It also enables diverse image manipulation including random jittering, image morphing, and category transfer. Such highly flexible restoration and manipulation are made possible through relaxing the assumption of existing GAN inversion methods, which tend to fix the generator. Notably, we allow the generator to be fine-tuned on-the-fly in a progressive manner regularized by feature distance obtained by the discriminator in GAN. We show that these easy-to-implement and practical changes help preserve the reconstruction to remain in the manifold of nature images, and thus lead to more precise and faithful reconstruction for real images. Code is available at https://github.com/XingangPan/deep-generative-prior.

We revisit human motion synthesis, a task useful in various real-world applications, in this paper. Whereas a number of methods have been developed previously for this task, they are often limited in two aspects: 1) focus on the poses while leaving the location movement behind, and 2) ignore the impact of the environment on the human motion. In this paper, we propose a new framework, with the interaction between the scene and the human motion taken into account. Considering the uncertainty of human motion, we formulate this task as a generative task, whose objective is to generate plausible human motion conditioned on both the scene and the human’s initial position. This framework factorizes the distribution of human motions into a distribution of movement trajectories conditioned on scenes and that of body pose dynamics conditioned on both scenes and trajectories. We further derive a GAN-based learning approach, with discriminators to enforce the compatibility between the human motion and the contextual scene as well as the 3D-to-2D projection constraints. We assess the effectiveness of the proposed method on two challenging datasets, which cover both synthetic and real-world environments.

Stereophonic audio, especially binaural audio, plays an essential role in immersive viewing environments. Recent research has explored generating visually guided stereophonic audios supervised by multi-channel audio collections. However, due to the requirement of professional recording devices, existing datasets are limited in scale and variety, which impedes the generalization of supervised methods in real-world scenarios. In this work, we propose PseudoBinaural, an effective pipeline that is free of binaural recordings. The key insight is to carefully build pseudo visual-stereo pairs with mono data for training. Specifically, we leverage spherical harmonic decomposition and head-related impulse response (HRIR) to identify the relationship between spatial locations and received binaural audios. Then in the visual modality, corresponding visual cues of the mono data are manually placed at sound source positions to form the pairs. Compared to fully-supervised paradigms, our binaural-recording-free pipeline shows great stability in cross-dataset evaluation and achieves comparable performance under subjective preference. Moreover, combined with binaural recordings, our method is able to further boost the performance of binaural audio generation under supervised settings1.

Most 3D shape completion approaches rely heavily on partial-complete shape pairs and learn in a fully super-vised manner. Despite their impressive performances on in-domain data, when generalizing to partial shapes in other forms or real-world partial scans, they often obtain unsatisfactory results due to domain gaps. In contrast to previous fully supervised approaches, in this paper we present ShapeInversion, which introduces Generative Adversarial Network (GAN) inversion to shape completion for the first time. ShapeInversion uses a GAN pre-trained on complete shapes by searching for a latent code that gives a complete shape that best reconstructs the given partial input. In this way, ShapeInversion no longer needs paired training data, and is capable of incorporating the rich prior captured in a well-trained generative model. On the ShapeNet bench-mark, the proposed ShapeInversion outperforms the SOTA unsupervised method, and is comparable with supervised methods that are learned using paired data. It also demonstrates remarkable generalization ability, giving robust results for real-world scans and partial inputs of various forms and incompleteness levels. Importantly, ShapeInversion naturally enables a series of additional abilities thanks to the involvement of a pre-trained GAN, such as producing multiple valid complete shapes for an ambiguous partial input, as well as shape manipulation and interpolation.

Image reconstruction and synthesis have witnessed remarkable progress thanks to the development of generative models. Nonetheless, gaps could still exist between the real and generated images, especially in the frequency domain. In this study, we show that narrowing gaps in the frequency domain can ameliorate image reconstruction and synthesis quality further. We propose a novel focal frequency loss, which allows a model to adaptively focus on frequency components that are hard to synthesize by down-weighting the easy ones. This objective function is complementary to existing spatial losses, offering great impedance against the loss of important frequency information due to the inherent bias of neural networks. We demonstrate the versatility and effectiveness of focal frequency loss to improve popular models, such as VAE, pix2pix, and SPADE, in both perceptual quality and quantitative performance. We further show its potential on StyleGAN2.1,2

City modeling is the foundation for computational urban planning, navigation, and entertainment. In this work, we present the first generative model of city blocks named BlockPlanner, and showcase its ability to synthesize valid city blocks with varying land lots configurations. We propose a novel vectorized city block representation utilizing a ring topology and a two-tier graph to capture the global and local structures of a city block. Each land lot is abstracted into a vector representation covering both its 3D geometry and land use semantics. Such vectorized representation enables us to deploy a lightweight network to capture the underlying distribution of land lots configurations in a city block. To enforce intrinsic spatial constraints of a valid city block, a set of effective loss functions are imposed to shape rational results. We contribute a pilot city block dataset to demonstrate the effectiveness and efficiency of our representation and framework over the state-of-the-art. Notably, our BlockPlanner is also able to edit and manipulate city blocks, enabling several useful applications, e.g., topology refinement and footprint generation.

Generative adversarial networks (GANs) typically require ample data for training in order to synthesize high-fidelity images. Recent studies have shown that training GANs with limited data remains formidable due to discriminator overfitting, the underlying cause that impedes the generator's convergence. This paper introduces a novel strategy called Adaptive Pseudo Augmentation (APA) to encourage healthy competition between the generator and the discriminator. As an alternative method to existing approaches that rely on standard data augmentations or model regularization, APA alleviates overfitting by employing the generator itself to augment the real data distribution with generated images, which deceives the discriminator adaptively. Extensive experiments demonstrate the effectiveness of APA in improving synthesis quality in the low-data regime. We provide a theoretical analysis to examine the convergence and rationality of our new training strategy. APA is simple and effective. It can be added seamlessly to powerful contemporary GANs, such as StyleGAN2, with negligible computational cost. Code: https://github.com/EndlessSora/DeceiveD.

The advancement of generative radiance fields has pushed the boundary of 3D-aware image synthesis. Motivated by the observation that a 3D object should look realistic from multiple viewpoints, these methods introduce a multi-view constraint as regularization to learn valid 3D radiance fields from 2D images. Despite the progress, they often fall short of capturing accurate 3D shapes due to the shape-color ambiguity, limiting their applicability in downstream tasks. In this work, we address this ambiguity by proposing a novel shading-guided generative implicit model that is able to learn a starkly improved shape representation. Our key insight is that an accurate 3D shape should also yield a realistic rendering under different lighting conditions. This multi-lighting constraint is realized by modeling illumination explicitly and performing shading with various lighting conditions. Gradients are derived by feeding the synthesized images to a discriminator. To compensate for the additional computational burden of calculating surface normals, we further devise an efficient volume rendering strategy via surface tracking, reducing the training and inference time by 24% and 48%, respectively. Our experiments on multiple datasets show that the proposed approach achieves photorealistic 3D-aware image synthesis while capturing accurate underlying 3D shapes. We demonstrate improved performance of our approach on 3D shape reconstruction against existing methods, and show its applicability on image relighting. Our code is available at https://github.com/XingangPan/ShadeGAN.

The advent of generative radiance fields has significantly promoted the development of 3D-aware image synthesis. The cumulative rendering process in radiance fields makes training these generative models much easier since gradients are distributed over the entire volume, but leads to diffused object surfaces. In the meantime, compared to radiance fields occupancy representations could inherently ensure deterministic surfaces. However, if we directly apply occupancy representations to generative models, during training they will only receive sparse gradients located on object surfaces and eventually suffer from the convergence problem. In this paper, we propose Generative Occupancy Fields (GOF), a novel model based on generative radiance fields that can learn compact object surfaces without impeding its training convergence. The key insight of GOF is a dedicated transition from the cumulative rendering in radiance fields to rendering with only the surface points as the learned surface gets more and more accurate. In this way, GOF combines the merits of two representations in a unified framework. In practice, the training-time transition of start from radiance fields and march to occupancy representations is achieved in GOF by gradually shrinking the sampling region in its rendering process from the entire volume to a minimal neighboring region around the surface. Through comprehensive experiments on multiple datasets, we demonstrate that GOF can synthesize high-quality images with 3D consistency and simultaneously learn compact and smooth object surfaces. Our code is available at https://github.com/SheldonTsui/GOF_NeurIPS2021.

Current methods for action recognition primarily rely on deep convolutional networks to derive feature embeddings of visual and motion features. While these methods have demonstrated remarkable performance on standard benchmarks, we are still in need of a better understanding as to how the videos, in particular their internal structures, relate to high-level semantics, which may lead to benefits in multiple aspects, e.g. interpretable predictions and even new methods that can take the recognition performances to a next level. Towards this goal, we construct TAPOS, a new dataset developed on sport videos with manual annotations of sub-actions, and conduct a study on temporal action parsing on top. Our study shows that a sport activity usually consists of multiple sub-actions and that the awareness of such temporal structures is beneficial to action recognition. We also investigate a number of temporal parsing methods, and thereon devise an improved method that is capable of mining sub-actions from training data without knowing the labels of them. On the constructed TAPOS, the proposed method is shown to reveal intra-action information, i.e. how action instances are made of sub-actions, and inter-action information, i.e. one specific sub-action may commonly appear in various actions.

On public benchmarks, current action recognition techniques have achieved great success. However, when used in real-world applications, e.g. sport analysis, which requires the capability of parsing an activity into phases and differentiating between subtly different actions, their performances remain far from being satisfactory. To take action recognition to a new level, we develop FineGym, a new dataset built on top of gymnasium videos. Compared to existing action recognition datasets, FineGym is distinguished in richness, quality, and diversity. In particular, it provides temporal annotations at both action and sub-action levels with a three-level semantic hierarchy. For example, a “balance beam” activity will be annotated as a sequence of elementary sub-actions derived from five sets: “leap-jump-hop”, “beam-turns”, “flight-salto”, “flight-handspring”, and “dismount”, where the sub-action in each set will be further annotated with finely defined class labels. This new level of granularity presents significant challenges for action recognition, e.g. how to parse the temporal structures from a coherent action, and how to distinguish between subtly different action classes. We systematically investigates different methods on this dataset and obtains a number of interesting findings. We hope this dataset could advance research towards action understanding.

Visual tempo characterizes the dynamics and the temporal scale of an action. Modeling such visual tempos of different actions facilitates their recognition. Previous works often capture the visual tempo through sampling raw videos at multiple rates and constructing an input-level frame pyramid, which usually requires a costly multi-branch network to handle. In this work we propose a generic Temporal Pyramid Network (TPN) at the feature-level, which can be flexibly integrated into 2D or 3D backbone networks in a plug-and-play manner. Two essential components of TPN, the source of features and the fusion of features, form a feature hierarchy for the backbone so that it can capture action instances at various tempos. TPN also shows consistent improvements over other challenging baselines on several action recognition datasets. Specifically, when equipped with TPN, the 3D ResNet-50 with dense sampling obtains a 2\% gain on the validation set of Kinetics-400. A further analysis also reveals that TPN gains most of its improvements on action classes that have large variances in their visual tempos, validating the effectiveness of TPN.

Natural scene understanding is a challenging task, particularly when encountering images of multiple objects that are partially occluded. This obstacle is given rise by varying object ordering and positioning. Existing scene understanding paradigms are able to parse only the visible parts, resulting in incomplete and unstructured scene interpretation. In this paper, we investigate the problem of scene de-occlusion, which aims to recover the underlying occlusion ordering and complete the invisible parts of occluded objects. We make the first attempt to address the problem through a novel and unified framework that recovers hidden scene structures without ordering and amodal annotations as supervisions. This is achieved via Partial Completion Network (PCNet)-mask (M) and -content (C), that learn to recover fractions of object masks and contents, respectively, in a self-supervised manner. Based on PCNet-M and PCNet-C, we devise a novel inference scheme to accomplish scene de-occlusion, via progressive ordering recovery, amodal completion and content completion. Extensive experiments on real-world scenes demonstrate the superior performance of our approach to other alternatives. Remarkably, our approach that is trained in a self-supervised manner achieves comparable results to fully-supervised methods. The proposed scene de-occlusion framework benefits many applications, including high-quality and controllable image manipulation and scene recomposition (see Fig. 1), as well as the conversion of existing modal mask annotations to amodal mask annotations.

Sounds provide rich semantics, complementary to visual data, for many tasks. However, in practice, sounds from multiple sources are often mixed together. In this paper we propose a novel framework, referred to as MinusPlus Network (MP-Net), for the task of visual sound separation. MP-Net separates sounds recursively in the order of average energy, removing the separated sound from the mixture at the end of each prediction, until the mixture becomes empty or contains only noise. In this way, MP-Net could be applied to sound mixtures with arbitrary numbers and types of sounds. Moreover, while MP-Net keeps removing sounds with large energy from the mixture, sounds with small energy could emerge and become clearer, so that the separation is more accurate. Compared to previous methods, MP-Net obtains state-of-the-art results on two large scale datasets, across mixtures with different types and numbers of sounds.

Mainstream captioning models often follow a sequential structure to generate cap- tions, leading to issues such as introduction of irrelevant semantics, lack of diversity in the generated captions, and inadequate generalization performance. In this paper, we present an alternative paradigm for image captioning, which factorizes the captioning procedure into two stages: (1) extracting an explicit semantic representation from the given image; and (2) constructing the caption based on a recursive compositional procedure in a bottom-up manner. Compared to conventional ones, our paradigm better preserves the semantic content through an explicit factorization of semantics and syntax. By using the compositional generation procedure, caption construction follows a recursive structure, which naturally fits the properties of human language. Moreover, the proposed compositional procedure requires less data to train, generalizes better, and yields more diverse captions.

Relationships among objects play a crucial role in image understanding. Despite the great success of deep learning techniques in recognizing individual objects, reasoning about the relationships among objects remains a challenging task. Previous methods often treat this as a classification problem, considering each type of relationship (e.g. ride) or each distinct visual phrase (e.g. person-ride-horse) as a category. Such approaches are faced with significant difficulties caused by the high diversity of visual appearance for each kind of relationships or the large number of distinct visual phrases. We propose an integrated framework to tackle this problem. At the heart of this framework is the Deep Relational Network, a novel formulation designed specifically for exploiting the statistical dependencies between objects and their relationships. On two large data sets, the proposed method achieves substantial improvement over state-of-the-art.

Despite the substantial progress in recent years, the image captioning techniques are still far from being perfect. Sentences produced by existing methods, e.g. those based on RNNs, are often overly rigid and lacking in variability. This issue is related to a learning principle widely used in practice, that is, to maximize the likelihood of training samples. This principle encourages high resemblance to the “ground-truth” captions, while suppressing other reasonable descriptions. Conventional evaluation metrics, e.g. BLEU and METEOR, also favor such restrictive methods. In this paper, we explore an alternative approach, with the aim to improve the naturalness and diversity - two essential properties of human expression. Specifically, we propose a new framework based on Conditional Generative Adversarial Networks (CGAN), which jointly learns a generator to produce descriptions conditioned on images and an evaluator to assess how well a description fits the visual content. It is noteworthy that training a sequence generator is nontrivial. We overcome the difficulty by Policy Gradient, a strategy stemming from Reinforcement Learning, which allows the generator to receive early feedback along the way. We tested our method on two large datasets, where it performed competitively against real people in our user study and outperformed other methods on various tasks.

Image captioning, a popular topic in computer vision, has achieved substantial progress in recent years. However, the distinctiveness of natural descriptions is often overlooked in previous work. It is closely related to the quality of captions, as distinctive captions are more likely to describe images with their unique aspects. In this work, we propose a new learning method, Contrastive Learning (CL), for image captioning. Specifically, via two constraints formulated on top of a reference model, the proposed method can encourage distinctiveness, while maintaining the overall quality of the generated captions. We tested our method on two challenging datasets, where it improves the baseline model by significant margins. We also showed in our studies that the proposed method is generic and can be used for models with various structures.