Siwei Zhang

Siwei Zhang, Qianli Ma, Yan Zhang et al.

Automatic perception of human behaviors during social interactions is crucial for AR/VR applications, and an essential component is estimation of plausible 3D human pose and shape of our social partners from the egocentric view. One of the biggest challenges of this task is severe body truncation due to close social distances in egocentric scenarios, which brings large pose ambiguities for unseen body parts. To tackle this challenge, we propose a novel scene-conditioned diffusion method to model the body pose distribution. Conditioned on the 3D scene geometry, the diffusion model generates bodies in plausible human-scene interactions, with the sampling guided by a physics-based collision score to further resolve human-scene inter-penetrations. The classifier-free training enables flexible sampling with different conditions and enhanced diversity. A visibility-aware graph convolution model guided by per-joint visibility serves as the diffusion denoiser to incorporate inter-joint dependencies and per-body-part control. Extensive evaluations show that our method generates bodies in plausible interactions with 3D scenes, achieving both superior accuracy for visible joints and diversity for invisible body parts. The code is available at https://sanweiliti.github.io/egohmr/egohmr.html.

Rui Wang, Sophokles Ktistakis, Siwei Zhang et al.

The surgical usage of Mixed Reality (MR) has received growing attention in areas such as surgical navigation systems, skill assessment, and robot-assisted surgeries. For such applications, pose estimation for hand and surgical instruments from an egocentric perspective is a fundamental task and has been studied extensively in the computer vision field in recent years. However, the development of this field has been impeded by a lack of datasets, especially in the surgical field, where bloody gloves and reflective metallic tools make it hard to obtain 3D pose annotations for hands and objects using conventional methods. To address this issue, we propose POV-Surgery, a large-scale, synthetic, egocentric dataset focusing on pose estimation for hands with different surgical gloves and three orthopedic surgical instruments, namely scalpel, friem, and diskplacer. Our dataset consists of 53 sequences and 88,329 frames, featuring high-resolution RGB-D video streams with activity annotations, accurate 3D and 2D annotations for hand-object pose, and 2D hand-object segmentation masks. We fine-tune the current SOTA methods on POV-Surgery and further show the generalizability when applying to real-life cases with surgical gloves and tools by extensive evaluations. The code and the dataset are publicly available at batfacewayne.github.io/POV_Surgery_io/.

Xi Chen, Yongxiang Liao, Yun Xiong et al.

Temporal Interaction Graphs (TIGs) are widely employed to model intricate real-world systems such as financial systems and social networks. To capture the dynamism and interdependencies of nodes, existing TIG embedding models need to process edges sequentially and chronologically. However, this requirement prevents it from being processed in parallel and struggle to accommodate burgeoning data volumes to GPU. Consequently, many large-scale temporal interaction graphs are confined to CPU processing. Furthermore, a generalized GPU scaling and acceleration approach remains unavailable. To facilitate large-scale TIGs' implementation on GPUs for acceleration, we introduce a novel training approach namely Streaming Edge Partitioning and Parallel Acceleration for Temporal Interaction Graph Embedding (SPEED). The SPEED is comprised of a Streaming Edge Partitioning Component (SEP) which addresses space overhead issue by assigning fewer nodes to each GPU, and a Parallel Acceleration Component (PAC) which enables simultaneous training of different sub-graphs, addressing time overhead issue. Our method can achieve a good balance in computing resources, computing time, and downstream task performance. Empirical validation across 7 real-world datasets demonstrates the potential to expedite training speeds by a factor of up to 19.29x. Simultaneously, resource consumption of a single-GPU can be diminished by up to 69%, thus enabling the multiple GPU-based training and acceleration encompassing millions of nodes and billions of edges. Furthermore, our approach also maintains its competitiveness in downstream tasks.

Siwei Zhang, Yun Xiong, Yao Zhang et al.

Continuous-time dynamic graph modeling is a crucial task for many real-world applications, such as financial risk management and fraud detection. Though existing dynamic graph modeling methods have achieved satisfactory results, they still suffer from three key limitations, hindering their scalability and further applicability. i) Indiscriminate updating. For incoming edges, existing methods would indiscriminately deal with them, which may lead to more time consumption and unexpected noisy information. ii) Ineffective node-wise long-term modeling. They heavily rely on recurrent neural networks (RNNs) as a backbone, which has been demonstrated to be incapable of fully capturing node-wise long-term dependencies in event sequences. iii) Neglect of re-occurrence patterns. Dynamic graphs involve the repeated occurrence of neighbors that indicates their importance, which is disappointedly neglected by existing methods. In this paper, we present iLoRE, a novel dynamic graph modeling method with instant node-wise Long-term modeling and Re-occurrence preservation. To overcome the indiscriminate updating issue, we introduce the Adaptive Short-term Updater module that will automatically discard the useless or noisy edges, ensuring iLoRE's effectiveness and instant ability. We further propose the Long-term Updater to realize more effective node-wise long-term modeling, where we innovatively propose the Identity Attention mechanism to empower a Transformer-based updater, bypassing the limited effectiveness of typical RNN-dominated designs. Finally, the crucial re-occurrence patterns are also encoded into a graph module for informative representation learning, which will further improve the expressiveness of our method. Our experimental results on real-world datasets demonstrate the effectiveness of our iLoRE for dynamic graph modeling.

Siwei Zhang, Yun Xiong, Yao Zhang et al.

Temporal Graph Networks (TGNs) have shown remarkable performance in learning representation for continuous-time dynamic graphs. However, real-world dynamic graphs typically contain diverse and intricate noise. Noise can significantly degrade the quality of representation generation, impeding the effectiveness of TGNs in downstream tasks. Though structure learning is widely applied to mitigate noise in static graphs, its adaptation to dynamic graph settings poses two significant challenges. i) Noise dynamics. Existing structure learning methods are ill-equipped to address the temporal aspect of noise, hampering their effectiveness in such dynamic and ever-changing noise patterns. ii) More severe noise. Noise may be introduced along with multiple interactions between two nodes, leading to the re-pollution of these nodes and consequently causing more severe noise compared to static graphs. In this paper, we present RDGSL, a representation learning method in continuous-time dynamic graphs. Meanwhile, we propose dynamic graph structure learning, a novel supervisory signal that empowers RDGSL with the ability to effectively combat noise in dynamic graphs. To address the noise dynamics issue, we introduce the Dynamic Graph Filter, where we innovatively propose a dynamic noise function that dynamically captures both current and historical noise, enabling us to assess the temporal aspect of noise and generate a denoised graph. We further propose the Temporal Embedding Learner to tackle the challenge of more severe noise, which utilizes an attention mechanism to selectively turn a blind eye to noisy edges and hence focus on normal edges, enhancing the expressiveness for representation generation that remains resilient to noise. Our method demonstrates robustness towards downstream tasks, resulting in up to 5.1% absolute AUC improvement in evolving classification versus the second-best baseline.

Yifeng Wang, Zhouhong Gu, Siwei Zhang et al.

Explainable fake news detection predicts the authenticity of news items with annotated explanations. Today, Large Language Models (LLMs) are known for their powerful natural language understanding and explanation generation abilities. However, presenting LLMs for explainable fake news detection remains two main challenges. Firstly, fake news appears reasonable and could easily mislead LLMs, leaving them unable to understand the complex news-faking process. Secondly, utilizing LLMs for this task would generate both correct and incorrect explanations, which necessitates abundant labor in the loop. In this paper, we propose LLM-GAN, a novel framework that utilizes prompting mechanisms to enable an LLM to become Generator and Detector and for realistic fake news generation and detection. Our results demonstrate LLM-GAN's effectiveness in both prediction performance and explanation quality. We further showcase the integration of LLM-GAN to a cloud-native AI platform to provide better fake news detection service in the cloud.

Xi Chen, Yun Xiong, Siwei Zhang et al.

Discrete-Time Dynamic Graphs (DTDGs), which are prevalent in real-world implementations and notable for their ease of data acquisition, have garnered considerable attention from both academic researchers and industry practitioners. The representation learning of DTDGs has been extensively applied to model the dynamics of temporally changing entities and their evolving connections. Currently, DTDG representation learning predominantly relies on GNN+RNN architectures, which manifest the inherent limitations of both Graph Neural Networks (GNNs) and Recurrent Neural Networks (RNNs). GNNs suffer from the over-smoothing issue as the models architecture goes deeper, while RNNs struggle to capture long-term dependencies effectively. GNN+RNN architectures also grapple with scaling to large graph sizes and long sequences. Additionally, these methods often compute node representations separately and focus solely on individual node characteristics, thereby overlooking the behavior intersections between the two nodes whose link is being predicted, such as instances where the two nodes appear together in the same context or share common neighbors. This paper introduces a novel representation learning method DTFormer for DTDGs, pivoting from the traditional GNN+RNN framework to a Transformer-based architecture. Our approach exploits the attention mechanism to concurrently process topological information within the graph at each timestamp and temporal dynamics of graphs along the timestamps, circumventing the aforementioned fundamental weakness of both GNNs and RNNs. Moreover, we enhance the model's expressive capability by incorporating the intersection relationships among nodes and integrating a multi-patching module. Extensive experiments conducted on six public dynamic graph benchmark datasets confirm our model's efficacy, achieving the SOTA performance.

Zhiyin Qian, Siwei Zhang, Bharat Lal Bhatnagar et al.

Human motion reconstruction from monocular videos is a fundamental challenge in computer vision, with broad applications in AR/VR, robotics, and digital content creation, but remains challenging under frequent occlusions in real-world settings. Existing regression-based methods are efficient but fragile to missing observations, while optimization- and diffusion-based approaches improve robustness at the cost of slow inference speed and heavy preprocessing steps. To address these limitations, we leverage recent advances in generative masked modeling and present MoRo: Masked Modeling for human motion Recovery under Occlusions. MoRo is an occlusion-robust, end-to-end generative framework that formulates motion reconstruction as a video-conditioned task, and efficiently recover human motion in a consistent global coordinate system from RGB videos. By masked modeling, MoRo naturally handles occlusions while enabling efficient, end-to-end inference. To overcome the scarcity of paired video-motion data, we design a cross-modality learning scheme that learns multi-modal priors from a set of heterogeneous datasets: (i) a trajectory-aware motion prior trained on MoCap datasets, (ii) an image-conditioned pose prior trained on image-pose datasets, capturing diverse per-frame poses, and (iii) a video-conditioned masked transformer that fuses motion and pose priors, finetuned on video-motion datasets to integrate visual cues with motion dynamics for robust inference. Extensive experiments on EgoBody and RICH demonstrate that MoRo substantially outperforms state-of-the-art methods in accuracy and motion realism under occlusions, while performing on-par in non-occluded scenarios. MoRo achieves real-time inference at 70 FPS on a single H200 GPU.

Siwei Zhang, Yun Xiong, Xi Chen et al.

Agentic Reinforcement Learning (Agentic RL) has shown remarkable potential in large language model-based (LLM) agents. These works can empower LLM agents to tackle complex tasks via multi-step, tool-integrated reasoning. However, an inherent limitation of existing Agentic RL methods is their reliance on a pure on-policy paradigm for exploration, restricting exploration to the agent's self-generated outputs and preventing the discovery of new reasoning perspectives for further improvement. While recent efforts incorporate auxiliary off-policy signals to enhance exploration, they typically utilize full off-policy trajectories for trajectory-level policy estimation, overlooking the necessity for the fine-grained, step-level exploratory dynamics within agentic rollout. In this paper, we revisit exploration in Agentic RL and propose Retrieval-Augmented Policy Optimization (RAPO), a novel RL framework that introduces retrieval to explicitly expand exploration during training. To achieve this, we decompose the Agentic RL training process into two phases: (i) Hybrid-policy Agentic Rollout, and (ii) Retrieval-aware Policy Optimization. Specifically, we propose a Hybrid-policy Agentic Rollout strategy, which allows the agents to continuously reason over the retrieved off-policy step-level traces. It dynamically extends the reasoning receptive field of agents, enabling broader exploration conditioned on external behaviors. Subsequently, we introduce the Retrieval-aware Policy Optimization mechanism, which calibrates the policy gradient estimation with retrieval reward and importance shaping, stabilizing training and prioritizing retrieval-illuminating exploration. Extensive experiments show that RAPO achieves an +5.0% average gain on fourteen datasets across three agentic reasoning tasks, while delivering 1.2x faster training efficiency.

Siwei Zhang, Bharat Lal Bhatnagar, Yuanlu Xu et al.

We propose RoHM, an approach for robust 3D human motion reconstruction from monocular RGB(-D) videos in the presence of noise and occlusions. Most previous approaches either train neural networks to directly regress motion in 3D or learn data-driven motion priors and combine them with optimization at test time. The former do not recover globally coherent motion and fail under occlusions; the latter are time-consuming, prone to local minima, and require manual tuning. To overcome these shortcomings, we exploit the iterative, denoising nature of diffusion models. RoHM is a novel diffusion-based motion model that, conditioned on noisy and occluded input data, reconstructs complete, plausible motions in consistent global coordinates. Given the complexity of the problem -- requiring one to address different tasks (denoising and infilling) in different solution spaces (local and global motion) -- we decompose it into two sub-tasks and learn two models, one for global trajectory and one for local motion. To capture the correlations between the two, we then introduce a novel conditioning module, combining it with an iterative inference scheme. We apply RoHM to a variety of tasks -- from motion reconstruction and denoising to spatial and temporal infilling. Extensive experiments on three popular datasets show that our method outperforms state-of-the-art approaches qualitatively and quantitatively, while being faster at test time. The code is available at https://sanweiliti.github.io/ROHM/ROHM.html.

Gen Li, Kaifeng Zhao, Siwei Zhang et al.

Understanding the world in first-person view is fundamental in Augmented Reality (AR). This immersive perspective brings dramatic visual changes and unique challenges compared to third-person views. Synthetic data has empowered third-person-view vision models, but its application to embodied egocentric perception tasks remains largely unexplored. A critical challenge lies in simulating natural human movements and behaviors that effectively steer the embodied cameras to capture a faithful egocentric representation of the 3D world. To address this challenge, we introduce EgoGen, a new synthetic data generator that can produce accurate and rich ground-truth training data for egocentric perception tasks. At the heart of EgoGen is a novel human motion synthesis model that directly leverages egocentric visual inputs of a virtual human to sense the 3D environment. Combined with collision-avoiding motion primitives and a two-stage reinforcement learning approach, our motion synthesis model offers a closed-loop solution where the embodied perception and movement of the virtual human are seamlessly coupled. Compared to previous works, our model eliminates the need for a pre-defined global path, and is directly applicable to dynamic environments. Combined with our easy-to-use and scalable data generation pipeline, we demonstrate EgoGen's efficacy in three tasks: mapping and localization for head-mounted cameras, egocentric camera tracking, and human mesh recovery from egocentric views. EgoGen will be fully open-sourced, offering a practical solution for creating realistic egocentric training data and aiming to serve as a useful tool for egocentric computer vision research. Refer to our project page: https://ego-gen.github.io/.

Siwei Zhang, Zhiwu Huang, Danda Pani Paudel et al.

Human emotions can be inferred from facial expressions. However, the annotations of facial expressions are often highly noisy in common emotion coding models, including categorical and dimensional ones. To reduce human labelling effort on multi-task labels, we introduce a new problem of facial emotion recognition with noisy multi-task annotations. For this new problem, we suggest a formulation from the point of joint distribution match view, which aims at learning more reliable correlations among raw facial images and multi-task labels, resulting in the reduction of noise influence. In our formulation, we exploit a new method to enable the emotion prediction and the joint distribution learning in a unified adversarial learning game. Evaluation throughout extensive experiments studies the real setups of the suggested new problem, as well as the clear superiority of the proposed method over the state-of-the-art competing methods on either the synthetic noisy labeled CIFAR-10 or practical noisy multi-task labeled RAF and AffectNet. The code is available at https://github.com/sanweiliti/noisyFER.

Zixuan Xiao, Jun Ma, Siwei Zhang

Understanding urban environment change is essential for sustainable development. However, current approaches, particularly remote sensing change detection, often rely on rigid, single-modal analysis. To overcome these limitations, we propose MMUEChange, a multi-modal agent framework that flexibly integrates heterogeneous urban data via a modular toolkit and a core module, Modality Controller for cross- and intra-modal alignment, enabling robust analysis of complex urban change scenarios. Case studies include: a shift toward small, community-focused parks in New York, reflecting local green space efforts; the spread of concentrated water pollution across districts in Hong Kong, pointing to coordinated water management; and a notable decline in open dumpsites in Shenzhen, with contrasting links between nighttime economic activity and waste types, indicating differing urban pressures behind domestic and construction waste. Compared to the best-performing baseline, the MMUEChange agent achieves a 46.7% improvement in task success rate and effectively mitigates hallucination, demonstrating its capacity to support complex urban change analysis tasks with real-world policy implications.

Xi Chen, Siwei Zhang, Yun Xiong et al.

Temporal Interaction Graphs (TIGs) are widely utilized to represent real-world systems. To facilitate representation learning on TIGs, researchers have proposed a series of TIG models. However, these models are still facing two tough gaps between the pre-training and downstream predictions in their ``pre-train, predict'' training paradigm. First, the temporal discrepancy between the pre-training and inference data severely undermines the models' applicability in distant future predictions on the dynamically evolving data. Second, the semantic divergence between pretext and downstream tasks hinders their practical applications, as they struggle to align with their learning and prediction capabilities across application scenarios. Recently, the ``pre-train, prompt'' paradigm has emerged as a lightweight mechanism for model generalization. Applying this paradigm is a potential solution to solve the aforementioned challenges. However, the adaptation of this paradigm to TIGs is not straightforward. The application of prompting in static graph contexts falls short in temporal settings due to a lack of consideration for time-sensitive dynamics and a deficiency in expressive power. To address this issue, we introduce Temporal Interaction Graph Prompting (TIGPrompt), a versatile framework that seamlessly integrates with TIG models, bridging both the temporal and semantic gaps. In detail, we propose a temporal prompt generator to offer temporally-aware prompts for different tasks. These prompts stand out for their minimalistic design, relying solely on the tuning of the prompt generator with very little supervision data. To cater to varying computational resource demands, we propose an extended ``pre-train, prompt-based fine-tune'' paradigm, offering greater flexibility. Through extensive experiments, the TIGPrompt demonstrates the SOTA performance and remarkable efficiency advantages.

Xi Chen, Yateng Tang, Jiarong Xu et al.

Effectively modeling time information and incorporating it into applications or models involving chronologically occurring events is crucial. Real-world scenarios often involve diverse and complex time patterns, which pose significant challenges for time encoding methods. While previous methods focus on capturing time patterns, many rely on specific inductive biases, such as using trigonometric functions to model periodicity. This narrow focus on single-pattern modeling makes them less effective in handling the diversity and complexities of real-world time patterns. In this paper, we investigate to improve the existing commonly used time encoding methods and introduce Learnable Transformation-based Generalized Time Encoding (LeTE). We propose using deep function learning techniques to parameterize non-linear transformations in time encoding, making them learnable and capable of modeling generalized time patterns, including diverse and complex temporal dynamics. By enabling learnable transformations, LeTE encompasses previous methods as specific cases and allows seamless integration into a wide range of tasks. Through extensive experiments across diverse domains, we demonstrate the versatility and effectiveness of LeTE.

Pengyiang Liu, Zhongyue Shi, Hongye Hao et al.

Video understanding requires models to continuously track and update world state during playback. While existing benchmarks have advanced video understanding evaluation across multiple dimensions, the observation of how models maintain world state remains insufficient. We propose VCBench, a streaming counting benchmark that repositions counting as a minimal probe for diagnosing world state maintenance capability. We decompose this capability into object counting (tracking currently visible objects vs.\ tracking cumulative unique identities) and event counting (detecting instantaneous actions vs.\ tracking complete activity cycles), forming 8 fine-grained subcategories. VCBench contains 406 videos with frame-by-frame annotations of 10,071 event occurrence moments and object state change moments, generating 1,000 streaming QA pairs with 4,576 query points along timelines. By observing state maintenance trajectories through streaming multi-point queries, we design three complementary metrics to diagnose numerical precision, trajectory consistency, and temporal awareness. Evaluation on mainstream video-language models shows that current models still exhibit significant deficiencies in spatial-temporal state maintenance, particularly struggling with tasks like periodic event counting. VCBench provides a diagnostic framework for measuring and improving state maintenance in video understanding systems.

Siwei Zhang, Yun Xiong, Yateng Tang et al.

Temporal graph neural networks (TGNNs) have shown remarkable performance in temporal graph modeling. However, real-world temporal graphs often possess rich textual information, giving rise to temporal text-attributed graphs (TTAGs). Such combination of dynamic text semantics and evolving graph structures introduces heightened complexity. Existing TGNNs embed texts statically and rely heavily on encoding mechanisms that biasedly prioritize structural information, overlooking the temporal evolution of text semantics and the essential interplay between semantics and structures for synergistic reinforcement. To tackle these issues, we present \textbf{CROSS}, a flexible framework that seamlessly extends existing TGNNs for TTAG modeling. CROSS is designed by decomposing the TTAG modeling process into two phases: (i) temporal semantics extraction; and (ii) semantic-structural information unification. The key idea is to advance the large language models (LLMs) to dynamically extract the temporal semantics in text space and then generate cohesive representations unifying both semantics and structures. Specifically, we propose a Temporal Semantics Extractor in the CROSS framework, which empowers LLMs to offer the temporal semantic understanding of node's evolving contexts of textual neighborhoods, facilitating semantic dynamics. Subsequently, we introduce the Semantic-structural Co-encoder, which collaborates with the above Extractor for synthesizing illuminating representations by jointly considering both semantic and structural information while encouraging their mutual reinforcement. Extensive experiments show that CROSS achieves state-of-the-art results on four public datasets and one industrial dataset, with 24.7% absolute MRR gain on average in temporal link prediction and 3.7% AUC gain in node classification of industrial application.

Evonne Ng, Siwei Zhang, Zhang Chen et al.

As embodied agents become central to VR, telepresence, and digital human applications, their motion must go beyond speech-aligned gestures: agents should turn toward users, respond to their movement, and maintain natural gaze. Current methods lack this spatial awareness. We close this gap with the first real-time, fully causal method for spatially-aware conversational motion, deployable on a streaming VR headset. Given a user's position and dyadic audio, our approach produces full-body motion that aligns gestures with speech while orienting the agent according to the user. Our architecture combines a causal transformer-based VAE with interleaved latent tokens for streaming inference and a flow matching model conditioned on user trajectory and audio. To support varying gaze preferences, we introduce a gaze scoring mechanism with classifier-free guidance to decouple learning from control: the model captures natural spatial alignment from data, while users can adjust eye contact intensity at inference time. On the Embody 3D dataset, our method achieves state-of-the-art motion quality at over 300 FPS -- 3x faster than non-causal baselines -- while capturing the subtle spatial dynamics of natural conversation. We validate our approach on a live VR system, bringing spatially-aware conversational agents to real-time deployment. Please see https://evonneng.github.io/sarah/ for details.

Claire McLean, Makenzie Meendering, Tristan Swartz et al.

The Codec Avatars Lab at Meta introduces Embody 3D, a multimodal dataset of 500 individual hours of 3D motion data from 439 participants collected in a multi-camera collection stage, amounting to over 54 million frames of tracked 3D motion. The dataset features a wide range of single-person motion data, including prompted motions, hand gestures, and locomotion; as well as multi-person behavioral and conversational data like discussions, conversations in different emotional states, collaborative activities, and co-living scenarios in an apartment-like space. We provide tracked human motion including hand tracking and body shape, text annotations, and a separate audio track for each participant.

Marko Mihajlovic, Siwei Zhang, Gen Li et al.

Parametric human body models play a crucial role in computer graphics and vision, enabling applications ranging from human motion analysis to understanding human-environment interactions. Traditionally, these models use surface meshes, which pose challenges in efficiently handling interactions with other geometric entities, such as objects and scenes, typically represented as meshes or point clouds. To address this limitation, recent research has explored volumetric neural implicit body models. However, existing works are either insufficiently robust for complex human articulations or impose high computational and memory costs, limiting their widespread use. To this end, we introduce VolumetricSMPL, a neural volumetric body model that leverages Neural Blend Weights (NBW) to generate compact, yet efficient MLP decoders. Unlike prior approaches that rely on large MLPs, NBW dynamically blends a small set of learned weight matrices using predicted shape- and pose-dependent coefficients, significantly improving computational efficiency while preserving expressiveness. VolumetricSMPL outperforms prior volumetric occupancy model COAP with 10x faster inference, 6x lower GPU memory usage, enhanced accuracy, and a Signed Distance Function (SDF) for efficient and differentiable contact modeling. We demonstrate VolumetricSMPL's strengths across four challenging tasks: (1) reconstructing human-object interactions from in-the-wild images, (2) recovering human meshes in 3D scenes from egocentric views, (3) scene-constrained motion synthesis, and (4) resolving self-intersections. Our results highlight its broad applicability and significant performance and efficiency gains.

Siwei Zhang, Xi Chen, Yun Xiong et al.

Temporal Graph Networks (TGNs) have demonstrated their remarkable performance in modeling temporal interaction graphs. These works can generate temporal node representations by encoding the surrounding neighborhoods for the target node. However, an inherent limitation of existing TGNs is their reliance on fixed, hand-crafted rules for neighborhood encoding, overlooking the necessity for an adaptive and learnable neighborhood that can accommodate both personalization and temporal evolution across different timestamps. In this paper, we aim to enhance existing TGNs by introducing an adaptive neighborhood encoding mechanism. We present SEAN, a flexible plug-and-play model that can be seamlessly integrated with existing TGNs, effectively boosting their performance. To achieve this, we decompose the adaptive neighborhood encoding process into two phases: (i) representative neighbor selection, and (ii) temporal-aware neighborhood information aggregation. Specifically, we propose the Representative Neighbor Selector component, which automatically pinpoints the most important neighbors for the target node. It offers a tailored understanding of each node's unique surrounding context, facilitating personalization. Subsequently, we propose a Temporal-aware Aggregator, which synthesizes neighborhood aggregation by selectively determining the utilization of aggregation routes and decaying the outdated information, allowing our model to adaptively leverage both the contextually significant and current information during aggregation. We conduct extensive experiments by integrating SEAN into three representative TGNs, evaluating their performance on four public datasets and one financial benchmark dataset introduced in this paper. The results demonstrate that SEAN consistently leads to performance improvements across all models, achieving SOTA performance and exceptional robustness.

Yan Wu, Jiahao Wang, Yan Zhang et al.

The synthesis of human grasping has numerous applications including AR/VR, video games and robotics. While methods have been proposed to generate realistic hand-object interaction for object grasping and manipulation, these typically only consider interacting hand alone. Our goal is to synthesize whole-body grasping motions. Starting from an arbitrary initial pose, we aim to generate diverse and natural whole-body human motions to approach and grasp a target object in 3D space. This task is challenging as it requires modeling both whole-body dynamics and dexterous finger movements. To this end, we propose SAGA (StochAstic whole-body Grasping with contAct), a framework which consists of two key components: (a) Static whole-body grasping pose generation. Specifically, we propose a multi-task generative model, to jointly learn static whole-body grasping poses and human-object contacts. (b) Grasping motion infilling. Given an initial pose and the generated whole-body grasping pose as the start and end of the motion respectively, we design a novel contact-aware generative motion infilling module to generate a diverse set of grasp-oriented motions. We demonstrate the effectiveness of our method, which is a novel generative framework to synthesize realistic and expressive whole-body motions that approach and grasp randomly placed unseen objects. Code and models are available at https://jiahaoplus.github.io/SAGA/saga.html.

Siwei Zhang, Qianli Ma, Yan Zhang et al.

Understanding social interactions from egocentric views is crucial for many applications, ranging from assistive robotics to AR/VR. Key to reasoning about interactions is to understand the body pose and motion of the interaction partner from the egocentric view. However, research in this area is severely hindered by the lack of datasets. Existing datasets are limited in terms of either size, capture/annotation modalities, ground-truth quality, or interaction diversity. We fill this gap by proposing EgoBody, a novel large-scale dataset for human pose, shape and motion estimation from egocentric views, during interactions in complex 3D scenes. We employ Microsoft HoloLens2 headsets to record rich egocentric data streams (including RGB, depth, eye gaze, head and hand tracking). To obtain accurate 3D ground truth, we calibrate the headset with a multi-Kinect rig and fit expressive SMPL-X body meshes to multi-view RGB-D frames, reconstructing 3D human shapes and poses relative to the scene, over time. We collect 125 sequences, spanning diverse interaction scenarios, and propose the first benchmark for 3D full-body pose and shape estimation of the social partner from egocentric views. We extensively evaluate state-of-the-art methods, highlight their limitations in the egocentric scenario, and address such limitations leveraging our high-quality annotations. Data and code are available at https://sanweiliti.github.io/egobody/egobody.html.

Siwei Zhang, Yan Zhang, Federica Bogo et al.

Recovering high-quality 3D human motion in complex scenes from monocular videos is important for many applications, ranging from AR/VR to robotics. However, capturing realistic human-scene interactions, while dealing with occlusions and partial views, is challenging; current approaches are still far from achieving compelling results. We address this problem by proposing LEMO: LEarning human MOtion priors for 4D human body capture. By leveraging the large-scale motion capture dataset AMASS, we introduce a novel motion smoothness prior, which strongly reduces the jitters exhibited by poses recovered over a sequence. Furthermore, to handle contacts and occlusions occurring frequently in body-scene interactions, we design a contact friction term and a contact-aware motion infiller obtained via per-instance self-supervised training. To prove the effectiveness of the proposed motion priors, we combine them into a novel pipeline for 4D human body capture in 3D scenes. With our pipeline, we demonstrate high-quality 4D human body capture, reconstructing smooth motions and physically plausible body-scene interactions. The code and data are available at https://sanweiliti.github.io/LEMO/LEMO.html.

Siwei Zhang, Yan Zhang, Qianli Ma et al.

High fidelity digital 3D environments have been proposed in recent years, however, it remains extremely challenging to automatically equip such environment with realistic human bodies. Existing work utilizes images, depth or semantic maps to represent the scene, and parametric human models to represent 3D bodies. While being straightforward, their generated human-scene interactions are often lack of naturalness and physical plausibility. Our key observation is that humans interact with the world through body-scene contact. To synthesize realistic human-scene interactions, it is essential to effectively represent the physical contact and proximity between the body and the world. To that end, we propose a novel interaction generation method, named PLACE (Proximity Learning of Articulation and Contact in 3D Environments), which explicitly models the proximity between the human body and the 3D scene around it. Specifically, given a set of basis points on a scene mesh, we leverage a conditional variational autoencoder to synthesize the minimum distances from the basis points to the human body surface. The generated proximal relationship exhibits which region of the scene is in contact with the person. Furthermore, based on such synthesized proximity, we are able to effectively obtain expressive 3D human bodies that interact with the 3D scene naturally. Our perceptual study shows that PLACE significantly improves the state-of-the-art method, approaching the realism of real human-scene interaction. We believe our method makes an important step towards the fully automatic synthesis of realistic 3D human bodies in 3D scenes. The code and model are available for research at https://sanweiliti.github.io/PLACE/PLACE.html.

Yan Wu, Aoming Liu, Zhiwu Huang et al.

This paper aims at enlarging the problem of Neural Architecture Search (NAS) from Single-Path and Multi-Path Search to automated Mixed-Path Search. In particular, we model the NAS problem as a sparse supernet using a new continuous architecture representation with a mixture of sparsity constraints. The sparse supernet enables us to automatically achieve sparsely-mixed paths upon a compact set of nodes. To optimize the proposed sparse supernet, we exploit a hierarchical accelerated proximal gradient algorithm within a bi-level optimization framework. Extensive experiments on Convolutional Neural Network and Recurrent Neural Network search demonstrate that the proposed method is capable of searching for compact, general and powerful neural architectures.

Yunxuan Zhang, Siwei Zhang, Yue He et al.

To enable realistic shape (e.g. pose and expression) transfer, existing face reenactment methods rely on a set of target faces for learning subject-specific traits. However, in real-world scenario end-users often only have one target face at hand, rendering existing methods inapplicable. In this work, we bridge this gap by proposing a novel one-shot face reenactment learning framework. Our key insight is that the one-shot learner should be able to disentangle and compose appearance and shape information for effective modeling. Specifically, the target face appearance and the source face shape are first projected into latent spaces with their corresponding encoders. Then these two latent spaces are associated by learning a shared decoder that aggregates multi-level features to produce the final reenactment results. To further improve the synthesizing quality on mustache and hair regions, we additionally propose FusionNet which combines the strengths of our learned decoder and the traditional warping method. Extensive experiments show that our one-shot face reenactment system achieves superior transfer fidelity as well as identity preserving capability than alternatives. More remarkably, our approach trained with only one target image per subject achieves competitive results to those using a set of target images, demonstrating the practical merit of this work. Code, models and an additional set of reenacted faces have been publicly released at the project page.