Cewu Lu

Sanqing Qu, Tianpei Zou, Florian Roehrbein et al.

Deep neural networks (DNNs) often perform poorly in the presence of domain shift and category shift. How to upcycle DNNs and adapt them to the target task remains an important open problem. Unsupervised Domain Adaptation (UDA), especially recently proposed Source-free Domain Adaptation (SFDA), has become a promising technology to address this issue. Nevertheless, existing SFDA methods require that the source domain and target domain share the same label space, consequently being only applicable to the vanilla closed-set setting. In this paper, we take one step further and explore the Source-free Universal Domain Adaptation (SF-UniDA). The goal is to identify "known" data samples under both domain and category shift, and reject those "unknown" data samples (not present in source classes), with only the knowledge from standard pre-trained source model. To this end, we introduce an innovative global and local clustering learning technique (GLC). Specifically, we design a novel, adaptive one-vs-all global clustering algorithm to achieve the distinction across different target classes and introduce a local k-NN clustering strategy to alleviate negative transfer. We examine the superiority of our GLC on multiple benchmarks with different category shift scenarios, including partial-set, open-set, and open-partial-set DA. Remarkably, in the most challenging open-partial-set DA scenario, GLC outperforms UMAD by 14.8\% on the VisDA benchmark. The code is available at https://github.com/ispc-lab/GLC.

Yifan Zhang, Bo Pang, Cewu Lu

Typical vision backbones manipulate structured features. As a compromise, semantic segmentation has long been modeled as per-point prediction on dense regular grids. In this work, we present a novel and efficient modeling that starts from interpreting the image as a tessellation of learnable regions, each of which has flexible geometrics and carries homogeneous semantics. To model region-wise context, we exploit Transformer to encode regions in a sequence-to-sequence manner by applying multi-layer self-attention on the region embeddings, which serve as proxies of specific regions. Semantic segmentation is now carried out as per-region prediction on top of the encoded region embeddings using a single linear classifier, where a decoder is no longer needed. The proposed RegProxy model discards the common Cartesian feature layout and operates purely at region level. Hence, it exhibits the most competitive performance-efficiency trade-off compared with the conventional dense prediction methods. For example, on ADE20K, the small-sized RegProxy-S/16 outperforms the best CNN model using 25% parameters and 4% computation, while the largest RegProxy-L/16 achieves 52.9mIoU which outperforms the state-of-the-art by 2.1% with fewer resources. Codes and models are available at https://github.com/YiF-Zhang/RegionProxy.

Wenqiang Xu, Zhenjun Yu, Han Xue et al.

Tactile sensing is one of the modalities humans rely on heavily to perceive the world. Working with vision, this modality refines local geometry structure, measures deformation at the contact area, and indicates the hand-object contact state. With the availability of open-source tactile sensors such as DIGIT, research on visual-tactile learning is becoming more accessible and reproducible. Leveraging this tactile sensor, we propose a novel visual-tactile in-hand object reconstruction framework \textbf{VTacO}, and extend it to \textbf{VTacOH} for hand-object reconstruction. Since our method can support both rigid and deformable object reconstruction, no existing benchmarks are proper for the goal. We propose a simulation environment, VT-Sim, which supports generating hand-object interaction for both rigid and deformable objects. With VT-Sim, we generate a large-scale training dataset and evaluate our method on it. Extensive experiments demonstrate that our proposed method can outperform the previous baseline methods qualitatively and quantitatively. Finally, we directly apply our model trained in simulation to various real-world test cases, which display qualitative results. Codes, models, simulation environment, and datasets are available at \url{https://sites.google.com/view/vtaco/}.

Xinyu Xu, Shengcheng Luo, Yanchao Yang et al.

Building a general-purpose intelligent home-assistant agent skilled in diverse tasks by human commands is a long-term blueprint of embodied AI research, which poses requirements on task planning, environment modeling, and object interaction. In this work, we study primitive mobile manipulations for embodied agents, i.e. how to navigate and interact based on an instructed verb-noun pair. We propose DISCO, which features non-trivial advancements in contextualized scene modeling and efficient controls. In particular, DISCO incorporates differentiable scene representations of rich semantics in object and affordance, which is dynamically learned on the fly and facilitates navigation planning. Besides, we propose dual-level coarse-to-fine action controls leveraging both global and local cues to accomplish mobile manipulation tasks efficiently. DISCO easily integrates into embodied tasks such as embodied instruction following. To validate our approach, we take the ALFRED benchmark of large-scale long-horizon vision-language navigation and interaction tasks as a test bed. In extensive experiments, we make comprehensive evaluations and demonstrate that DISCO outperforms the art by a sizable +8.6% success rate margin in unseen scenes, even without step-by-step instructions. Our code is publicly released at https://github.com/AllenXuuu/DISCO.

Yong-Lu Li, Yue Xu, Xinyu Xu et al.

Understanding objects is a central building block of artificial intelligence, especially for embodied AI. Even though object recognition excels with deep learning, current machines still struggle to learn higher-level knowledge, e.g., what attributes an object has, and what can we do with an object. In this work, we propose a challenging Object Concept Learning (OCL) task to push the envelope of object understanding. It requires machines to reason out object affordances and simultaneously give the reason: what attributes make an object possesses these affordances. To support OCL, we build a densely annotated knowledge base including extensive labels for three levels of object concept (category, attribute, affordance), and the causal relations of three levels. By analyzing the causal structure of OCL, we present a baseline, Object Concept Reasoning Network (OCRN). It leverages causal intervention and concept instantiation to infer the three levels following their causal relations. In experiments, OCRN effectively infers the object knowledge while following the causalities well. Our data and code are available at https://mvig-rhos.com/ocl.

Jiefeng Li, Ye Yuan, Davis Rempe et al.

Estimating global human motion from moving cameras is challenging due to the entanglement of human and camera motions. To mitigate the ambiguity, existing methods leverage learned human motion priors, which however often result in oversmoothed motions with misaligned 2D projections. To tackle this problem, we propose COIN, a control-inpainting motion diffusion prior that enables fine-grained control to disentangle human and camera motions. Although pre-trained motion diffusion models encode rich motion priors, we find it non-trivial to leverage such knowledge to guide global motion estimation from RGB videos. COIN introduces a novel control-inpainting score distillation sampling method to ensure well-aligned, consistent, and high-quality motion from the diffusion prior within a joint optimization framework. Furthermore, we introduce a new human-scene relation loss to alleviate the scale ambiguity by enforcing consistency among the humans, camera, and scene. Experiments on three challenging benchmarks demonstrate the effectiveness of COIN, which outperforms the state-of-the-art methods in terms of global human motion estimation and camera motion estimation. As an illustrative example, COIN outperforms the state-of-the-art method by 33% in world joint position error (W-MPJPE) on the RICH dataset.

Jun Lv, Yunhai Feng, Cheng Zhang et al.

Model-based reinforcement learning (MBRL) is recognized with the potential to be significantly more sample-efficient than model-free RL. How an accurate model can be developed automatically and efficiently from raw sensory inputs (such as images), especially for complex environments and tasks, is a challenging problem that hinders the broad application of MBRL in the real world. In this work, we propose a sensing-aware model-based reinforcement learning system called SAM-RL. Leveraging the differentiable physics-based simulation and rendering, SAM-RL automatically updates the model by comparing rendered images with real raw images and produces the policy efficiently. With the sensing-aware learning pipeline, SAM-RL allows a robot to select an informative viewpoint to monitor the task process. We apply our framework to real world experiments for accomplishing three manipulation tasks: robotic assembly, tool manipulation, and deformable object manipulation. We demonstrate the effectiveness of SAM-RL via extensive experiments. Videos are available on our project webpage at https://sites.google.com/view/rss-sam-rl.

Zhijie Deng, Jiaxin Shi, Hao Zhang et al.

This paper introduces a structured, adaptive-length deep representation called Neural Eigenmap. Unlike prior spectral methods such as Laplacian Eigenmap that operate in a nonparametric manner, Neural Eigenmap leverages NeuralEF to parametrically model eigenfunctions using a neural network. We show that, when the eigenfunction is derived from positive relations in a data augmentation setup, applying NeuralEF results in an objective function that resembles those of popular self-supervised learning methods, with an additional symmetry-breaking property that leads to \emph{structured} representations where features are ordered by importance. We demonstrate using such representations as adaptive-length codes in image retrieval systems. By truncation according to feature importance, our method requires up to $16\times$ shorter representation length than leading self-supervised learning ones to achieve similar retrieval performance. We further apply our method to graph data and report strong results on a node representation learning benchmark with more than one million nodes.

Mingyu Zhang, Jiting Cai, Mingyu Liu et al.

Visual reasoning, as a prominent research area, plays a crucial role in AI by facilitating concept formation and interaction with the world. However, current works are usually carried out separately on small datasets thus lacking generalization ability. Through rigorous evaluation of diverse benchmarks, we demonstrate the shortcomings of existing ad-hoc methods in achieving cross-domain reasoning and their tendency to data bias fitting. In this paper, we revisit visual reasoning with a two-stage perspective: (1) symbolization and (2) logical reasoning given symbols or their representations. We find that the reasoning stage is better at generalization than symbolization. Thus, it is more efficient to implement symbolization via separated encoders for different data domains while using a shared reasoner. Given our findings, we establish design principles for visual reasoning frameworks following the separated symbolization and shared reasoning. The proposed two-stage framework achieves impressive generalization ability on various visual reasoning tasks, including puzzles, physical prediction, and visual question answering (VQA), encompassing both 2D and 3D modalities. We believe our insights will pave the way for generalizable visual reasoning.

Ying Feng, Hongjie Fang, Yinong He et al.

Dexterous robotic hands enable robots to perform complex manipulations that require fine-grained control and adaptability. Achieving such manipulation is challenging because the high degrees of freedom tightly couple hand and arm motions, making learning and control difficult. Successful dexterous manipulation relies not only on precise hand motions, but also on accurate spatial positioning of the arm and coordinated arm-hand dynamics. However, most existing visuomotor policies represent arm and hand actions in a single combined space, which often causes high-dimensional hand actions to dominate the coupled action space and compromise arm control. To address this, we propose DQ-RISE, which quantizes hand states to simplify hand motion prediction while preserving essential patterns, and applies a continuous relaxation that allows arm actions to diffuse jointly with these compact hand states. This design enables the policy to learn arm-hand coordination from data while preventing hand actions from overwhelming the action space. Experiments show that DQ-RISE achieves more balanced and efficient learning, paving the way toward structured and generalizable dexterous manipulation. Project website: http://rise-policy.github.io/DQ-RISE/

Weikun Peng, Jun Lv, Yuwei Zeng et al.

The tie-knotting task is highly challenging due to the tie's high deformation and long-horizon manipulation actions. This work presents TieBot, a Real-to-Sim-to-Real learning from visual demonstration system for the robots to learn to knot a tie. We introduce the Hierarchical Feature Matching approach to estimate a sequence of tie's meshes from the demonstration video. With these estimated meshes used as subgoals, we first learn a teacher policy using privileged information. Then, we learn a student policy with point cloud observation by imitating teacher policy. Lastly, our pipeline applies learned policy to real-world execution. We demonstrate the effectiveness of TieBot in simulation and the real world. In the real-world experiment, a dual-arm robot successfully knots a tie, achieving 50% success rate among 10 trials. Videos can be found https://tiebots.github.io/.

Xin Li, Siyuan Huang, Qiaojun Yu et al.

Automating garment manipulation poses a significant challenge for assistive robotics due to the diverse and deformable nature of garments. Traditional approaches typically require separate models for each garment type, which limits scalability and adaptability. In contrast, this paper presents a unified approach using vision-language models (VLMs) to improve keypoint prediction across various garment categories. By interpreting both visual and semantic information, our model enables robots to manage different garment states with a single model. We created a large-scale synthetic dataset using advanced simulation techniques, allowing scalable training without extensive real-world data. Experimental results indicate that the VLM-based method significantly enhances keypoint detection accuracy and task success rates, providing a more flexible and general solution for robotic garment manipulation. In addition, this research also underscores the potential of VLMs to unify various garment manipulation tasks within a single framework, paving the way for broader applications in home automation and assistive robotics for future.

Sanqing Qu, Tianpei Zou, Florian Röhrbein et al.

Deep neural networks often exhibit sub-optimal performance under covariate and category shifts. Source-Free Domain Adaptation (SFDA) presents a promising solution to this dilemma, yet most SFDA approaches are restricted to closed-set scenarios. In this paper, we explore Source-Free Universal Domain Adaptation (SF-UniDA) aiming to accurately classify "known" data belonging to common categories and segregate them from target-private "unknown" data. We propose a novel Global and Local Clustering (GLC) technique, which comprises an adaptive one-vs-all global clustering algorithm to discern between target classes, complemented by a local k-NN clustering strategy to mitigate negative transfer. Despite the effectiveness, the inherent closed-set source architecture leads to uniform treatment of "unknown" data, impeding the identification of distinct "unknown" categories. To address this, we evolve GLC to GLC++, integrating a contrastive affinity learning strategy. We examine the superiority of GLC and GLC++ across multiple benchmarks and category shift scenarios. Remarkably, in the most challenging open-partial-set scenarios, GLC and GLC++ surpass GATE by 16.8\% and 18.9\% in H-score on VisDA, respectively. GLC++ enhances the novel category clustering accuracy of GLC by 4.1\% in open-set scenarios on Office-Home. Furthermore, the introduced contrastive learning strategy not only enhances GLC but also significantly facilitates existing methodologies. The code is available at https://github.com/ispc-lab/GLC-plus.

Xiaoyang Liu, Boran Wen, Xinpeng Liu et al.

Spatio-temporal Human-Object Interaction (ST-HOI) understanding aims at detecting HOIs from videos, which is crucial for activity understanding. However, existing whole-body-object interaction video benchmarks overlook the truth that open-world objects are diverse, that is, they usually provide limited and predefined object classes. Therefore, we introduce a new open-world benchmark: Grounding Interacted Objects (GIO) including 1,098 interacted objects class and 290K interacted object boxes annotation. Accordingly, an object grounding task is proposed expecting vision systems to discover interacted objects. Even though today's detectors and grounding methods have succeeded greatly, they perform unsatisfactorily in localizing diverse and rare objects in GIO. This profoundly reveals the limitations of current vision systems and poses a great challenge. Thus, we explore leveraging spatio-temporal cues to address object grounding and propose a 4D question-answering framework (4D-QA) to discover interacted objects from diverse videos. Our method demonstrates significant superiority in extensive experiments compared to current baselines. Data and code will be publicly available at https://github.com/DirtyHarryLYL/HAKE-AVA.

Yang Jin, Jun Lv, Wenye Yu et al.

Self-improvement requires robotic systems to initially learn from human-provided data and then gradually enhance their capabilities through interaction with the environment. This is similar to how humans improve their skills through continuous practice. However, achieving effective self-improvement is challenging, primarily because robots tend to repeat their existing abilities during interactions, often failing to generate new, valuable data for learning. In this paper, we identify the key to successful self-improvement: modal-level exploration and data selection. By incorporating a modal-level exploration mechanism during policy execution, the robot can produce more diverse and multi-modal interactions. At the same time, we select the most valuable trials and high-quality segments from these interactions for learning. We successfully demonstrate effective robot self-improvement on both simulation benchmarks and real-world experiments. The capability for self-improvement will enable us to develop more robust and high-success-rate robotic control strategies at a lower cost. Our code and experiment scripts are available at https://ericjin2002.github.io/SIME/

Wendi Chen, Han Xue, Yi Wang et al.

Human-level contact-rich manipulation relies on the distinct roles of two key modalities: vision provides spatially rich but temporally slow global context, while force sensing captures rapid, high-frequency local contact dynamics. Integrating these signals is challenging due to their fundamental frequency and informational disparities. In this work, we propose ImplicitRDP, a unified end-to-end visual-force diffusion policy that integrates visual planning and reactive force control within a single network. We introduce Structural Slow-Fast Learning, a mechanism utilizing causal attention to simultaneously process asynchronous visual and force tokens, allowing the policy to perform closed-loop adjustments at the force frequency while maintaining the temporal coherence of action chunks. Furthermore, to mitigate modality collapse where end-to-end models fail to adjust the weights across different modalities, we propose Virtual-target-based Representation Regularization. This auxiliary objective maps force feedback into the same space as the action, providing a stronger, physics-grounded learning signal than raw force prediction. Extensive experiments on contact-rich tasks demonstrate that ImplicitRDP significantly outperforms both vision-only and hierarchical baselines, achieving superior reactivity and success rates with a streamlined training pipeline. Code and videos will be publicly available at https://implicit-rdp.github.io.

Chenxi Wang, Hao-Shu Fang, Minghao Gou et al.

Efficient and robust grasp pose detection is vital for robotic manipulation. For general 6 DoF grasping, conventional methods treat all points in a scene equally and usually adopt uniform sampling to select grasp candidates. However, we discover that ignoring where to grasp greatly harms the speed and accuracy of current grasp pose detection methods. In this paper, we propose "graspness", a quality based on geometry cues that distinguishes graspable areas in cluttered scenes. A look-ahead searching method is proposed for measuring the graspness and statistical results justify the rationality of our method. To quickly detect graspness in practice, we develop a neural network named cascaded graspness model to approximate the searching process. Extensive experiments verify the stability, generality and effectiveness of our graspness model, allowing it to be used as a plug-and-play module for different methods. A large improvement in accuracy is witnessed for various previous methods after equipping our graspness model. Moreover, we develop GSNet, an end-to-end network that incorporates our graspness model for early filtering of low-quality predictions. Experiments on a large-scale benchmark, GraspNet-1Billion, show that our method outperforms previous arts by a large margin (30+ AP) and achieves a high inference speed. The library of GSNet has been integrated into AnyGrasp, which is at https://github.com/graspnet/anygrasp_sdk.

Yue Xu, Zhilin Lin, Yusong Qiu et al.

Though dataset distillation has witnessed rapid development in recent years, the distillation of multimodal data, e.g., image-text pairs, poses unique and under-explored challenges. Unlike unimodal data, image-text contrastive learning (ITC) data lack inherent categorization and should instead place greater emphasis on modality correspondence. In this work, we propose Low-Rank Similarity Mining (LoRS) for multimodal dataset distillation, that concurrently distills a ground truth similarity matrix with image-text pairs, and leverages low-rank factorization for efficiency and scalability. The proposed approach brings significant improvement to the existing algorithms, marking a significant contribution to the field of visual-language dataset distillation. We advocate adopting LoRS as a foundational synthetic data setup for image-text dataset distillation. Our code is available at https://github.com/silicx/LoRS_Distill.

Xinyu Xu, Yong-Lu Li, Cewu Lu

Anticipating future events is an essential feature for intelligent systems and embodied AI. However, compared to the traditional recognition task, the uncertainty of future and reasoning ability requirement make the anticipation task very challenging and far beyond solved. In this filed, previous methods usually care more about the model architecture design or but few attention has been put on how to train an anticipation model with a proper learning policy. To this end, in this work, we propose a novel training scheme called Dynamic Context Removal (DCR), which dynamically schedules the visibility of observed future in the learning procedure. It follows the human-like curriculum learning process, i.e., gradually removing the event context to increase the anticipation difficulty till satisfying the final anticipation target. Our learning scheme is plug-and-play and easy to integrate any reasoning model including transformer and LSTM, with advantages in both effectiveness and efficiency. In extensive experiments, the proposed method achieves state-of-the-art on four widely-used benchmarks. Our code and models are publicly released at https://github.com/AllenXuuu/DCR.

Yong-Lu Li, Yue Xu, Xinyu Xu et al.

Attributes and objects can compose diverse compositions. To model the compositional nature of these concepts, it is a good choice to learn them as transformations, e.g., coupling and decoupling. However, complex transformations need to satisfy specific principles to guarantee rationality. Here, we first propose a previously ignored principle of attribute-object transformation: Symmetry. For example, coupling peeled-apple with attribute peeled should result in peeled-apple, and decoupling peeled from apple should still output apple. Incorporating the symmetry, we propose a transformation framework inspired by group theory, i.e., SymNet. It consists of two modules: Coupling Network and Decoupling Network. We adopt deep neural networks to implement SymNet and train it in an end-to-end paradigm with the group axioms and symmetry as objectives. Then, we propose a Relative Moving Distance (RMD) based method to utilize the attribute change instead of the attribute pattern itself to classify attributes. Besides the compositions of single-attribute and object, our RMD is also suitable for complex compositions of multiple attributes and objects when incorporating attribute correlations. SymNet can be utilized for attribute learning, compositional zero-shot learning and outperforms the state-of-the-art on four widely-used benchmarks. Code is at https://github.com/DirtyHarryLYL/SymNet.

Kailin Li, Lixin Yang, Xinyu Zhan et al.

Estimating the articulated 3D hand-object pose from a single RGB image is a highly ambiguous and challenging problem, requiring large-scale datasets that contain diverse hand poses, object types, and camera viewpoints. Most real-world datasets lack these diversities. In contrast, data synthesis can easily ensure those diversities separately. However, constructing both valid and diverse hand-object interactions and efficiently learning from the vast synthetic data is still challenging. To address the above issues, we propose ArtiBoost, a lightweight online data enhancement method. ArtiBoost can cover diverse hand-object poses and camera viewpoints through sampling in a Composited hand-object Configuration and Viewpoint space (CCV-space) and can adaptively enrich the current hard-discernable items by loss-feedback and sample re-weighting. ArtiBoost alternatively performs data exploration and synthesis within a learning pipeline, and those synthetic data are blended into real-world source data for training. We apply ArtiBoost on a simple learning baseline network and witness the performance boost on several hand-object benchmarks. Our models and code are available at https://github.com/lixiny/ArtiBoost.

Bo Pang, Gao Peng, Yizhuo Li et al.

Convolutional video models have an order of magnitude larger computational complexity than their counterpart image-level models. Constrained by computational resources, there is no model or training method that can train long video sequences end-to-end. Currently, the main-stream method is to split a raw video into clips, leading to incomplete fragmentary temporal information flow. Inspired by natural language processing techniques dealing with long sentences, we propose to treat videos as serial fragments satisfying Markov property, and train it as a whole by progressively propagating information through the temporal dimension in multiple steps. This progressive training (PGT) method is able to train long videos end-to-end with limited resources and ensures the effective transmission of information. As a general and robust training method, we empirically demonstrate that it yields significant performance improvements on different models and datasets. As an illustrative example, the proposed method improves SlowOnly network by 3.7 mAP on Charades and 1.9 top-1 accuracy on Kinetics with negligible parameter and computation overhead. Code is available at https://github.com/BoPang1996/PGT.

Xinqi Zhu, Chang Xu, Langwen Hui et al.

We consider two less-emphasized temporal properties of video: 1. Temporal cues are fine-grained; 2. Temporal modeling needs reasoning. To tackle both problems at once, we exploit approximated bilinear modules (ABMs) for temporal modeling. There are two main points making the modules effective: two-layer MLPs can be seen as a constraint approximation of bilinear operations, thus can be used to construct deep ABMs in existing CNNs while reusing pretrained parameters; frame features can be divided into static and dynamic parts because of visual repetition in adjacent frames, which enables temporal modeling to be more efficient. Multiple ABM variants and implementations are investigated, from high performance to high efficiency. Specifically, we show how two-layer subnets in CNNs can be converted to temporal bilinear modules by adding an auxiliary-branch. Besides, we introduce snippet sampling and shifting inference to boost sparse-frame video classification performance. Extensive ablation studies are conducted to show the effectiveness of proposed techniques. Our models can outperform most state-of-the-art methods on Something-Something v1 and v2 datasets without Kinetics pretraining, and are also competitive on other YouTube-like action recognition datasets. Our code is available on https://github.com/zhuxinqimac/abm-pytorch.

Jiajun Tang, Jin Xia, Xinzhi Mu et al.

Understanding interaction is an essential part of video action detection. We propose the Asynchronous Interaction Aggregation network (AIA) that leverages different interactions to boost action detection. There are two key designs in it: one is the Interaction Aggregation structure (IA) adopting a uniform paradigm to model and integrate multiple types of interaction; the other is the Asynchronous Memory Update algorithm (AMU) that enables us to achieve better performance by modeling very long-term interaction dynamically without huge computation cost. We provide empirical evidence to show that our network can gain notable accuracy from the integrative interactions and is easy to train end-to-end. Our method reports the new state-of-the-art performance on AVA dataset, with 3.7 mAP gain (12.6% relative improvement) on validation split comparing to our strong baseline. The results on dataset UCF101-24 and EPIC-Kitchens further illustrate the effectiveness of our approach. Source code will be made public at: https://github.com/MVIG-SJTU/AlphAction .

Yong-Lu Li, Yue Xu, Xiaohan Mao et al.

Attributes and objects can compose diverse compositions. To model the compositional nature of these general concepts, it is a good choice to learn them through transformations, such as coupling and decoupling. However, complex transformations need to satisfy specific principles to guarantee the rationality. In this paper, we first propose a previously ignored principle of attribute-object transformation: Symmetry. For example, coupling peeled-apple with attribute peeled should result in peeled-apple, and decoupling peeled from apple should still output apple. Incorporating the symmetry principle, a transformation framework inspired by group theory is built, i.e. SymNet. SymNet consists of two modules, Coupling Network and Decoupling Network. With the group axioms and symmetry property as objectives, we adopt Deep Neural Networks to implement SymNet and train it in an end-to-end paradigm. Moreover, we propose a Relative Moving Distance (RMD) based recognition method to utilize the attribute change instead of the attribute pattern itself to classify attributes. Our symmetry learning can be utilized for the Compositional Zero-Shot Learning task and outperforms the state-of-the-art on widely-used benchmarks. Code is available at https://github.com/DirtyHarryLYL/SymNet.

Bo Pang, Kaiwen Zha, Hanwen Cao et al.

Extracting temporal and representation features efficiently plays a pivotal role in understanding visual sequence information. To deal with this, we propose a new recurrent neural framework that can be stacked deep effectively. There are mainly two novel designs in our deep RNN framework: one is a new RNN module called Context Bridge Module (CBM) which splits the information flowing along the sequence (temporal direction) and along depth (spatial representation direction), making it easier to train when building deep by balancing these two directions; the other is the Overlap Coherence Training Scheme that reduces the training complexity for long visual sequential tasks on account of the limitation of computing resources. We provide empirical evidence to show that our deep RNN framework is easy to optimize and can gain accuracy from the increased depth on several visual sequence problems. On these tasks, we evaluate our deep RNN framework with 15 layers, 7* than conventional RNN networks, but it is still easy to train. Our deep framework achieves more than 11% relative improvements over shallow RNN models on Kinetics, UCF-101, and HMDB-51 for video classification. For auxiliary annotation, after replacing the shallow RNN part of Polygon-RNN with our 15-layer deep CBM, the performance improves by 14.7%. For video future prediction, our deep RNN improves the state-of-the-art shallow model's performance by 2.4% on PSNR and SSIM. The code and trained models are published accompanied by this paper: https://github.com/BoPang1996/Deep-RNN-Framework.

Jiawen Yu, Hairuo Liu, Qiaojun Yu et al.

Vision-Language-Action (VLA) models have advanced general-purpose robotic manipulation by leveraging pretrained visual and linguistic representations. However, they struggle with contact-rich tasks that require fine-grained control involving force, especially under visual occlusion or dynamic uncertainty. To address these limitations, we propose ForceVLA, a novel end-to-end manipulation framework that treats external force sensing as a first-class modality within VLA systems. ForceVLA introduces FVLMoE, a force-aware Mixture-of-Experts fusion module that dynamically integrates pretrained visual-language embeddings with real-time 6-axis force feedback during action decoding. This enables context-aware routing across modality-specific experts, enhancing the robot's ability to adapt to subtle contact dynamics. We also introduce \textbf{ForceVLA-Data}, a new dataset comprising synchronized vision, proprioception, and force-torque signals across five contact-rich manipulation tasks. ForceVLA improves average task success by 23.2% over strong pi_0-based baselines, achieving up to 80% success in tasks such as plug insertion. Our approach highlights the importance of multimodal integration for dexterous manipulation and sets a new benchmark for physically intelligent robotic control. Code and data will be released at https://sites.google.com/view/forcevla2025.

Wendi Chen, Han Xue, Fangyuan Zhou et al.

In recent years, imitation learning has made progress in the field of robotic manipulation. However, it still faces challenges when addressing complex long-horizon tasks with deformable objects, such as high-dimensional state spaces, complex dynamics, and multimodal action distributions. Traditional imitation learning methods often require a large amount of data and encounter distributional shifts and accumulative errors in these tasks. To address these issues, we propose a data-efficient general learning framework (DeformPAM) based on preference learning and reward-guided action selection. DeformPAM decomposes long-horizon tasks into multiple action primitives, utilizes 3D point cloud inputs and diffusion models to model action distributions, and trains an implicit reward model using human preference data. During the inference phase, the reward model scores multiple candidate actions, selecting the optimal action for execution, thereby reducing the occurrence of anomalous actions and improving task completion quality. Experiments conducted on three challenging real-world long-horizon deformable object manipulation tasks demonstrate the effectiveness of this method. Results show that DeformPAM improves both task completion quality and efficiency compared to baseline methods even with limited data. Code and data will be available at https://deform-pam.robotflow.ai.

Jiacheng Liu, Pengxiang Ding, Qihang Zhou et al.

Recent Vision-Language-Action models show potential to generalize across embodiments but struggle to quickly align with a new robot's action space when high-quality demonstrations are scarce, especially for bipedal humanoids. We present TrajBooster, a cross-embodiment framework that leverages abundant wheeled-humanoid data to boost bipedal VLA. Our key idea is to use end-effector trajectories as a morphology-agnostic interface. TrajBooster (i) extracts 6D dual-arm end-effector trajectories from real-world wheeled humanoids, (ii) retargets them in simulation to Unitree G1 with a whole-body controller trained via a heuristic-enhanced harmonized online DAgger to lift low-dimensional trajectory references into feasible high-dimensional whole-body actions, and (iii) forms heterogeneous triplets that couple source vision/language with target humanoid-compatible actions to post-pre-train a VLA, followed by only 10 minutes of teleoperation data collection on the target humanoid domain. Deployed on Unitree G1, our policy achieves beyond-tabletop household tasks, enabling squatting, cross-height manipulation, and coordinated whole-body motion with markedly improved robustness and generalization. Results show that TrajBooster allows existing wheeled-humanoid data to efficiently strengthen bipedal humanoid VLA performance, reducing reliance on costly same-embodiment data while enhancing action space understanding and zero-shot skill transfer capabilities. For more details, For more details, please refer to our \href{https://jiachengliu3.github.io/TrajBooster/}.

Wenjing Tang, Xinyu He, Yongxi Huang et al.

Task planning under uncertainty is essential for home-service robots operating in the real world. Tasks involve ambiguous human instructions, hidden or unknown object locations, and open-vocabulary object types, leading to significant open-ended uncertainty and a boundlessly large planning space. To address these challenges, we propose Tru-POMDP, a planner that combines structured belief generation using Large Language Models (LLMs) with principled POMDP planning. Tru-POMDP introduces a hierarchical Tree of Hypotheses (TOH), which systematically queries an LLM to construct high-quality particle beliefs over possible world states and human goals. We further formulate an open-ended POMDP model that enables rigorous Bayesian belief tracking and efficient belief-space planning over these LLM-generated hypotheses. Experiments on complex object rearrangement tasks across diverse kitchen environments show that Tru-POMDP significantly outperforms state-of-the-art LLM-based and LLM-tree-search hybrid planners, achieving higher success rates with significantly better plans, stronger robustness to ambiguity and occlusion, and greater planning efficiency.

Zixuan Chen, Jiaxin Li, Liming Tan et al.

Intelligent robots need to interact with diverse objects across various environments. The appearance and state of objects frequently undergo complex transformations depending on the object properties, e.g., phase transitions. However, in the vision community, segmenting dynamic objects with phase transitions is overlooked. In light of this, we introduce the concept of phase in segmentation, which categorizes real-world objects based on their visual characteristics and potential morphological and appearance changes. Then, we present a new benchmark, Multi-Phase, Multi-Transition, and Multi-Scenery Video Object Segmentation (M$^3$-VOS), to verify the ability of models to understand object phases, which consists of 479 high-resolution videos spanning over 10 distinct everyday scenarios. It provides dense instance mask annotations that capture both object phases and their transitions. We evaluate state-of-the-art methods on M$^3$-VOS, yielding several key insights. Notably, current appearance-based approaches show significant room for improvement when handling objects with phase transitions. The inherent changes in disorder suggest that the predictive performance of the forward entropy-increasing process can be improved through a reverse entropy-reducing process. These findings lead us to propose ReVOS, a new plug-andplay model that improves its performance by reversal refinement. Our data and code will be publicly available at https://zixuan-chen.github.io/M-cube-VOS.github.io/.

Shengcheng Luo, Quanquan Peng, Jun Lv et al.

Employing a teleoperation system for gathering demonstrations offers the potential for more efficient learning of robot manipulation. However, teleoperating a robot arm equipped with a dexterous hand or gripper, via a teleoperation system presents inherent challenges due to the task's high dimensionality, complexity of motion, and differences between physiological structures. In this study, we introduce a novel system for joint learning between human operators and robots, that enables human operators to share control of a robot end-effector with a learned assistive agent, simplifies the data collection process, and facilitates simultaneous human demonstration collection and robot manipulation training. As data accumulates, the assistive agent gradually learns. Consequently, less human effort and attention are required, enhancing the efficiency of the data collection process. It also allows the human operator to adjust the control ratio to achieve a trade-off between manual and automated control. We conducted experiments in both simulated environments and physical real-world settings. Through user studies and quantitative evaluations, it is evident that the proposed system could enhance data collection efficiency and reduce the need for human adaptation while ensuring the collected data is of sufficient quality for downstream tasks. \textit{For more details, please refer to our webpage https://norweig1an.github.io/HAJL.github.io/.

Haoyuan Fu, Wenqiang Xu, Ruolin Ye et al.

Multiphysics phenomena, the coupling effects involving different aspects of physics laws, are pervasive in the real world and can often be encountered when performing everyday household tasks. Intelligent agents which seek to assist or replace human laborers will need to learn to cope with such phenomena in household task settings. To equip the agents with such kind of abilities, the research community needs a simulation environment, which will have the capability to serve as the testbed for the training process of these intelligent agents, to have the ability to support multiphysics coupling effects. Though many mature simulation software for multiphysics simulation have been adopted in industrial production, such techniques have not been applied to robot learning or embodied AI research. To bridge the gap, we propose a novel simulation environment named RFUniverse. This simulator can not only compute rigid and multi-body dynamics, but also multiphysics coupling effects commonly observed in daily life, such as air-solid interaction, fluid-solid interaction, and heat transfer. Because of the unique multiphysics capacities of this simulator, we can benchmark tasks that involve complex dynamics due to multiphysics coupling effects in a simulation environment before deploying to the real world. RFUniverse provides multiple interfaces to let the users interact with the virtual world in various ways, which is helpful and essential for learning, planning, and control. We benchmark three tasks with reinforcement learning, including food cutting, water pushing, and towel catching. We also evaluate butter pushing with a classic planning-control paradigm. This simulator offers an enhancement of physics simulation in terms of the computation of multiphysics coupling effects.

Liang Xu, Cuiling Lan, Wenjun Zeng et al.

Skeleton data carries valuable motion information and is widely explored in human action recognition. However, not only the motion information but also the interaction with the environment provides discriminative cues to recognize the action of persons. In this paper, we propose a joint learning framework for mutually assisted "interacted object localization" and "human action recognition" based on skeleton data. The two tasks are serialized together and collaborate to promote each other, where preliminary action type derived from skeleton alone helps improve interacted object localization, which in turn provides valuable cues for the final human action recognition. Besides, we explore the temporal consistency of interacted object as constraint to better localize the interacted object with the absence of ground-truth labels. Extensive experiments on the datasets of SYSU-3D, NTU60 RGB+D, Northwestern-UCLA and UAV-Human show that our method achieves the best or competitive performance with the state-of-the-art methods for human action recognition. Visualization results show that our method can also provide reasonable interacted object localization results.

Tutian Tang, Wenqiang Xu, Ruolin Ye et al.

Direct contour regression for instance segmentation is a challenging task. Previous works usually achieve it by learning to progressively refine the contour prediction or adopting a shape representation with limited expressiveness. In this work, we argue that the difficulty in regressing the contour points in one pass is mainly due to the ambiguity when discretizing a smooth contour into a polygon. To address the ambiguity, we propose a novel differentiable rendering-based approach named \textbf{ContourRender}. During training, it first predicts a contour generated by an invertible shape signature, and then optimizes the contour with the more stable silhouette by converting it to a contour mesh and rendering the mesh to a 2D map. This method significantly improves the quality of contour without iterations or cascaded refinements. Moreover, as optimization is not needed during inference, the inference speed will not be influenced. Experiments show the proposed ContourRender outperforms all the contour-based instance segmentation approaches on COCO, while stays competitive with the iteration-based state-of-the-art on Cityscapes. In addition, we specifically select a subset from COCO val2017 named COCO ContourHard-val to further demonstrate the contour quality improvements. Codes, models, and dataset split will be released.

Yunyan Hong, Ailing Zeng, Min Li et al.

Video action recognition (VAR) is a primary task of video understanding, and untrimmed videos are more common in real-life scenes. Untrimmed videos have redundant and diverse clips containing contextual information, so sampling dense clips is essential. Recently, some works attempt to train a generic model to select the N most representative clips. However, it is difficult to model the complex relations from intra-class clips and inter-class videos within a single model and fixed selected number, and the entanglement of multiple relations is also hard to explain. Thus, instead of "only look once", we argue "divide and conquer" strategy will be more suitable in untrimmed VAR. Inspired by the speed reading mechanism, we propose a simple yet effective clip-level solution based on skim-scan techniques. Specifically, the proposed Skim-Scan framework first skims the entire video and drops those uninformative and misleading clips. For the remaining clips, it scans clips with diverse features gradually to drop redundant clips but cover essential content. The above strategies can adaptively select the necessary clips according to the difficulty of the different videos. To trade off the computational complexity and performance, we observe the similar statistical expression between lightweight and heavy networks, thus it supports us to explore the combination of them. Comprehensive experiments are performed on ActivityNet and mini-FCVID datasets, and results demonstrate that our solution surpasses the state-of-the-art performance in terms of both accuracy and efficiency.

Lixin Yang, Jiasen Li, Wenqiang Xu et al.

3D hand estimation has been a long-standing research topic in computer vision. A recent trend aims not only to estimate the 3D hand joint locations but also to recover the mesh model. However, achieving those goals from a single RGB image remains challenging. In this paper, we introduce an end-to-end learnable model, BiHand, which consists of three cascaded stages, namely 2D seeding stage, 3D lifting stage, and mesh generation stage. At the output of BiHand, the full hand mesh will be recovered using the joint rotations and shape parameters predicted from the network. Inside each stage, BiHand adopts a novel bisecting design which allows the networks to encapsulate two closely related information (e.g. 2D keypoints and silhouette in 2D seeding stage, 3D joints, and depth map in 3D lifting stage, joint rotations and shape parameters in the mesh generation stage) in a single forward pass. As the information represents different geometry or structure details, bisecting the data flow can facilitate optimization and increase robustness. For quantitative evaluation, we conduct experiments on two public benchmarks, namely the Rendered Hand Dataset (RHD) and the Stereo Hand Pose Tracking Benchmark (STB). Extensive experiments show that our model can achieve superior accuracy in comparison with state-of-the-art methods, and can produce appealing 3D hand meshes in several severe conditions.

Bo Pang, Kaiwen Zha, Hanwen Cao et al.

Understanding sequential information is a fundamental task for artificial intelligence. Current neural networks attempt to learn spatial and temporal information as a whole, limited their abilities to represent large scale spatial representations over long-range sequences. Here, we introduce a new modeling strategy called Semi-Coupled Structure (SCS), which consists of deep neural networks that decouple the complex spatial and temporal concepts learning. Semi-Coupled Structure can learn to implicitly separate input information into independent parts and process these parts respectively. Experiments demonstrate that a Semi-Coupled Structure can successfully annotate the outline of an object in images sequentially and perform video action recognition. For sequence-to-sequence problems, a Semi-Coupled Structure can predict future meteorological radar echo images based on observed images. Taken together, our results demonstrate that a Semi-Coupled Structure has the capacity to improve the performance of LSTM-like models on large scale sequential tasks.

Hao-Shu Fang, Chenxi Wang, Minghao Gou et al.

Object grasping is critical for many applications, which is also a challenging computer vision problem. However, for the clustered scene, current researches suffer from the problems of insufficient training data and the lacking of evaluation benchmarks. In this work, we contribute a large-scale grasp pose detection dataset with a unified evaluation system. Our dataset contains 87,040 RGBD images with over 370 million grasp poses. Meanwhile, our evaluation system directly reports whether a grasping is successful or not by analytic computation, which is able to evaluate any kind of grasp poses without exhausted labeling pose ground-truth. We conduct extensive experiments to show that our dataset and evaluation system can align well with real-world experiments. Our dataset, source code and models will be made publicly available.

Jin Xia, Jiajun Tang, Cewu Lu

We present our three branch solutions for International Challenge on Activity Recognition at CVPR2019. This model seeks to fuse richer information of global video clip, short human attention and long-term human activity into a unified model. We have participated in two tasks: Task A, the Kinetics challenge and Task B, spatio-temporal action localization challenge. For Kinetics, we achieve 21.59% error rate. For the AVA challenge, our final model obtains 32.49% mAP on the test sets, which outperforms all submissions to the AVA challenge at CVPR 2018 for more than 10% mAP. As the future work, we will introduce human activity knowledge, which is a new dataset including key information of human activity.

Bo Pang, Kaiwen Zha, Cewu Lu

We introduce the first benchmark for a new problem --- recognizing human action adverbs (HAA): "Adverbs Describing Human Actions" (ADHA). This is the first step for computer vision to change over from pattern recognition to real AI. We demonstrate some key features of ADHA: a semantically complete set of adverbs describing human actions, a set of common, describable human actions, and an exhaustive labeling of simultaneously emerging actions in each video. We commit an in-depth analysis on the implementation of current effective models in action recognition and image captioning on adverb recognition, and the results show that such methods are unsatisfactory. Moreover, we propose a novel three-stream hybrid model to deal the HAA problem, which achieves a better result.

Wenqiang Xu, Yonglu Li, Cewu Lu

Instance segmentation is a problem of significance in computer vision. However, preparing annotated data for this task is extremely time-consuming and costly. By combining the advantages of 3D scanning, reasoning, and GAN-based domain adaptation techniques, we introduce a novel pipeline named SRDA to obtain large quantities of training samples with very minor effort. Our pipeline is well-suited to scenes that can be scanned, i.e. most indoor and some outdoor scenarios. To evaluate our performance, we build three representative scenes and a new dataset, with 3D models of various common objects categories and annotated real-world scene images. Extensive experiments show that our pipeline can achieve decent instance segmentation performance given very low human labor cost.

Cewu Lu, Hao Chen, Qifeng Chen et al.

The Imagenet Large Scale Visual Recognition Challenge (ILSVRC) is the one of the most important big data challenges to date. We participated in the object detection track of ILSVRC 2014 and received the fourth place among the 38 teams. We introduce in our object detection system a number of novel techniques in localization and recognition. For localization, initial candidate proposals are generated using selective search, and a novel bounding boxes regression method is used for better object localization. For recognition, to represent a candidate proposal, we adopt three features, namely, RCNN feature, IFV feature, and DPM feature. Given these features, category-specific combination functions are learned to improve the object recognition rate. In addition, object context in the form of background priors and object interaction priors are learned and applied in our system. Our ILSVRC 2014 results are reported alongside with the results of other participating teams.