※ 1) Useful Collections

• Github Recent LLM-based CV and related works.
• Github Must-read Papers on Large Language Model(LLM) Agents.
• Github Robotic Grasping Papers and Codes - Grasp Detection
• Github CV & Geometry-based 6DOF Robotic Grasping - 6D Grasp Pose Detection
• Github Diffusion-Literature-for-Robotics - Summary of key papers and blogs
• Github Awesome-Touch - Tactile Sensor and Simulator; Visual Tactile Manipulation; Open Source.
• Github awesome-embodied-vla/va/vln [vision-language-action (VLA), vision-language-navigation (VLN), vision-action (VA)]
• Github 🔥RSS2025 & CVPR2025 & ICLR2025 Embodied AI Paper List Resources
• Github my_arXiv_daily - Robotics

※ 2) Representative Blogs

• website (CCF) 具身智能 | CCF专家谈术语 (Cewu Lu)
• website (GraspNet) GraspNet通用物体抓取(GraspNet-1Billion + AnyGrasp + SuctionNet-1Billion + TransCG)
• website (AgiBot-World) 智元机器人联合上海人工智能实验室/国家地方共建人形机器人创新中心/上海库帕思，开源百万真机数据集

※ 3) Simulator/Robotics Toolkits

• Pangolin [This is a lightweight portable rapid development library for managing OpenGL display / interaction and abstracting video input.][github/stevenlovegrove][github/uoip]
• NLopt [This is a free/open-source library for nonlinear optimization, providing a common interface for different free optimization routines.][website/readthedocs]
• DART [DART (Dense Articulated Real-time Tracking) is a C++ library for tracking arbitrary articulated models with an RGB-D camera.][github/tschmidt23]
• 👍👍cuRobo [(arxiv2023.10) cuRobo: Parallelized Collision-Free Minimum-Jerk Robot Motion Generation] [arxiv link][project link][code|official][NVIDIA; Dieter Fox][CUDA Accelerated Robot Library]
• Gensim Gensim is a Python library for topic modelling, document indexing and similarity retrieval with large corpora. [github]
• Gym Gym is a standard API for reinforcement learning, and a diverse collection of reference environments [github]
• 👍👍Gymnasium Gymnasium is an API standard for reinforcement learning with a diverse collection of reference environments [github]
• Gymnasium-Robotics Gymnasium-Robotics is a collection of robotics simulation environments for Reinforcement Learning [github]
• 👍ManiSkill SAPIEN Manipulation Skill Framework, a GPU parallelized robotics simulator and benchmark [ManiSkill readthedocs]
• 👍PyRep PyRep is a toolkit for robot learning research, built on top of CoppeliaSim (previously called V-REP).
• CoppeliaSim It supports testing complex robotics systems via algorithms prototyping, kinematic design and digital twin creation.
• Deoxys A modular, real-time controller library for Franka Emika Panda robots, aiming to facilitate a wide range of robot learning research.
• ManiSkill Research ManiSkill helps propel groundbreaking research in generalizable robotic manipulation.
• Hillbot Scaling Robot Foundation Models via Simulation

• MetaWorld(CoRL2019)(arxiv2019.10) Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learning [paper link][arxiv link][project link][baseline method][code|official][Stanford University + UC Berkeley + Columbia University + University of Southern California + Robotics at Google; Chelsea Finn + Sergey Levine]

• 👍RLBench(RAL2020)(arxiv2019.09) RLBench: The Robot Learning Benchmark & Learning Environment [paper link][arxiv link][project link][code|official][Dyson Robotics Lab, Imperial College London][This dataset is based on Coppliasim 4.1.0 and PyRep]

• robosuite(2020.09) robosuite: A Modular Simulation Framework and Benchmark for Robot Learning [white paper][arxiv link][project link][documentation link][github link][robosuite.ai; robosuite is a simulation framework powered by the MuJoCo physics engine for robot learning. It also offers a suite of benchmark environments for reproducible research.]

• Ravens(TransporterNets)(CoRL2020)(arxiv2020.10) Transporter Networks: Rearranging the Visual World for Robotic Manipulation [paper link][arxiv link][project link][code|official][Robotics at Google][It trained robotic agents to learn pick and place with deep learning for vision-based manipulation in PyBullet.]

• 👍CALVIN(RAL2022)(Best Paper Award)(arxiv2021.12) Calvin: A Benchmark for Language-conditioned Policy Learning for Long-horizon Robot Manipulation Tasks [paper link][arxiv link][project link][code|official][University of Freiburg, Germany]

• VLMbench(NIPS2022 Datasets and Benchmarks)(arxiv2022.06) VLMbench: A Compositional Benchmark for Vision-and-Language Manipulation [paper link][arxiv link][project link][code|official][University of California + University of Michigan, It proposed the baseline method named 6D-CLIPort][This dataset is based on Coppliasim 4.1.0 and PyRep]

• Language-Table(RAL2023)(arxiv2022.10) Interactive Language: Talking to Robots in Real Time [paper link][arxiv link][project link][code|official][Robotics at Google][It is a suite of human-collected datasets and a multi-task continuous control benchmark for open vocabulary visuolinguomotor learning.]

• 👍ARNOLD(ICCV2023)(arxiv2023.04) ARNOLD: A Benchmark for Language-Grounded Task Learning with Continuous States in Realistic 3D Scenes [paper link][arxiv link][project link][code|official][dataset|official][challenges|official][UCLA + PKU + THU + Columbia University + BIGAI]

• RH20T(CoRLW2023)(arxiv2023.07) RH20T: A Comprehensive Robotic Dataset for Learning Diverse Skills in One-Shot [paper link][arxiv link][project link][SJTU][Its 150 skills were either selected from RLBench and MetaWorld, or proposed by themselves.]

• 👍BridgeData-V2(CoRL2023)(arxiv2023.08) BridgeData V2: A Dataset for Robot Learning at Scale [openreview link][paper link][arxiv link][project link][code|official][UC Berkeley + Stanford + Google DeepMind + CMU][It is based on the (arxiv2021.09) Bridge data: Boosting generalization of robotic skills with cross-domain datasets with [arxiv link] and [project link]]

• LoHoRavens(arxiv2023.10) LoHoRavens: A Long-Horizon Language-Conditioned Benchmark for Robotic Tabletop Manipulation [arxiv link][project link][code|official][LMU Munich + TUM][The code is largely based on method CLIPort-batchify(CoRL2021)(arxiv2021.09) and dataset Ravens(TransporterNets)(CoRL2020)]

• RoboHive(NIPS2023)(arxiv2023.10) RoboHive: A Unified Framework for Robot Learning [openreview link][paper link][arxiv link][project link][code|official][U.Washington + UC Berkeley + CMU + UT Austin + OpenAI + GoogleAI + Meta-AI; Datasets and Benchmarks Track]

• Safety-Gymnasium(NIPS2023 Datasets and Benchmarks)(arxiv2023.10) Safety-Gymnasium: A Unified Safe Reinforcement Learning Benchmark [openreview link][paper link][arxiv link][project link][code|official][PKU, Safety-Gymnasium is a highly scalable and customizable Safe Reinforcement Learning (SafeRL) library.]

• 👍Open X-Embodiment(RT-2-X)(arxiv2023.10) Open X-Embodiment: Robotic Learning Datasets and RT-X Models [arxiv link][project link][code|official][by Google DeepMind]

• 👍DROID(arxiv2024.03) DROID: A Large-Scale In-the-Wild Robot Manipulation Dataset [arxiv link][project link][dataset visualizer][code|official][Stanford + Berkeley + Toyota and many other universities; It used the diffusion policy for policy learning]

• 👍👍SimplerEnv(arxiv2024.05) Evaluating Real-World Robot Manipulation Policies in Simulation [arxiv link][project link][code|official][UC San Diego + Stanford University + UC Berkeley + Google DeepMind][Evaluating and reproducing real-world robot manipulation policies (e.g., RT-1, RT-1-X, Octo) in simulation under common setups (e.g., Google Robot, WidowX+Bridge)]

• 👍PerAct2(arxiv2024.07) PerAct2: Benchmarking and Learning for Robotic Bimanual Manipulation Tasks [arxiv link][project link][code|official][dataset link][University of Washington; Dieter Fox][This work extends previous work PerAct as well as RLBench for bimanual manipulation tasks.]

• 👍RoboTwin(ECCV Workshop 2024 Best Paper)(arxiv2024.09) RoboTwin: Dual-Arm Robot Benchmark with Generative Digital Twins [arxiv link][project link][code|official][The University of Hong Kong + AgileX Robotics + Shanghai AI Laboratory + Shenzhen University + Institute of Automation, Chinese Academy of Sciences; Ping Luo][AgileX Robotics (松灵机器人)]

• GEMBench(arxiv2024.10) Towards Generalizable Vision-Language Robotic Manipulation: A Benchmark and LLM-guided 3D Policy [arxiv link][project link][code|official][CNRS, PSL Research University, Shizhe Chen][It is still based on the RLBench]

• LADEV(arxiv2024.10) LADEV: A Language-Driven Testing and Evaluation Platform for Vision-Language-Action Models in Robotic Manipulation [arxiv link][project link][University of Alberta, Edmonto + University of Tokyo]

• EvalTasks(arxiv2024.10) On the Evaluation of Generative Robotic Simulations [arxiv link][project link][The University of Hong Kong + Tsinghua University IIIS + Shanghai Qi Zhi Institute + Shanghai AI Lab; Yi Ma + Huazhe Xu]

• 👍Synthetica(arxiv2024.10) Synthetica: Large Scale Synthetic Data for Robot Perception [arxiv link][project link][NVIDIA + University of Toronto]

• Mimicking-Bench(arxiv2024.12) Mimicking-Bench: A Benchmark for Generalizable Humanoid-Scene Interaction Learning via Human Mimicking [arxiv link][project link][Tsinghua University + Galbot + Shanghai Qi Zhi Institute + Shanghai Artificial Intelligence Laboratory + Peking University; He Wang]

• EMMOE(arxiv2025.03) EMMOE: A Comprehensive Benchmark for Embodied Mobile Manipulation in Open Environments [arxiv link][project link][code|official][Zhejiang University + University of Illinois Urbana-Champaign + University of Washington]

• 👍AgiBot-World(year2025.03) AgiBot World Colosseo: A Large-scale Manipulation Platform for Scalable and Intelligent Embodied Systems [pdf link][project link][dataset link][research blog][code|official][Team AgiBot-World]

• 👍**★-Gen(arxiv2025.03)** A Taxonomy for Evaluating Generalist Robot Policies [arxiv link][project link][Stanford University + Google DeepMind Robotics]

• 👍RoboVerse(year2025.04) RoboVerse: Towards a Unified Platform, Dataset and Benchmark for Scalable and Generalizable Robot Learning [paper link][project link][code|official][UC Berkeley + PKU + USC + UMich + UIUC + Stanford + CMU + UCLA + BIGAI]

• WorldEval(arxiv2025.05) WorldEval: World Model as Real-World Robot Policies Evaluator [arxiv link][project link][code|official][Midea Group + East China Normal University]

• AnyBody(arxiv2025.05) AnyBody: A Benchmark Suite for Cross-Embodiment Manipulation [arxiv link][project link][Princeton University; Jia Deng]

• Robo2VLM(arxiv2025.05) Robo2VLM: Visual Question Answering from Large-Scale In-the-Wild Robot Manipulation Datasets [arxiv link][code|official][University of California, Berkeley]

• RoboCulture(arxiv2025.05) RoboCulture: A Robotics Platform for Automated Biological Experimentation [arxiv link][project link][code|official][University of Toronto + Vector Institute + Toronto General Health Research Institute + Acceleration Consortium + Canadian Institute for Advanced Research + NVIDIA]

• AutoBio(arxiv2025.05) AutoBio: A Simulation and Benchmark for Robotic Automation in Digital Biology Laboratory [arxiv link][code|official][HKU + TeleAI + THU + SJTU + HKU + Shanghai Intelligent Computing Center; Xiaokang Yang + Xuelong Li + Ping Luo]

• DORI-Benchmark(arxiv2025.05) Right Side Up? Disentangling Orientation Understanding in MLLMs with Fine-grained Multi-axis Perception Tasks [arxiv link][dataset link][Boston University + Runway][DORI (Discriminative Orientation Reasoning Intelligence)]

• RoboCerebra(arxiv2025.06) RoboCerebra: A Large-scale Benchmark for Long-horizon Robotic Manipulation Evaluation [arxiv link][project link][dataset link][Beihang University + National University of Singapore + Shanghai Jiao Tong University]

• GenManip(arxiv2025.06) GENMANIP: LLM-driven Simulation for Generalizable Instruction-Following Manipulation [arxiv link][project link][code|official][Shanghai AI Laboratory + Xi'an Jiaotong University + Zhejiang University + Nanjing University; Jiangmiao Pang]

• CheckManual(CVPR2025, Highlight)(arxiv2025.06) CheckManual: A New Challenge and Benchmark for Manual-based Appliance Manipulation [paper link][arxiv link][project link][code|official][Peking University + PKU-Agibot Lab; Hao Dong]

• RoboArena(arxiv2025.06) RoboArena: Distributed Real-World Evaluation of Generalist Robot Policies [arxiv link][project link][code|official][University of California, Berkeley + Stanford University + University of Washington + Université de Montréal + NVIDIA + University of Pennsylvania + UT Austin + Yonsei University]

• 👍Survey(arxiv2023.12) Toward General-Purpose Robots via Foundation Models: A Survey and Meta-Analysis [arxiv link][code|official][Survey Paper of foundation models for robotics]

• 👍Survey(IJRR2024)(arxiv2023.12) Foundation Models in Robotics: Applications, Challenges, and the Future [paper link][arxiv link][code|official][Awesome-Robotics-Foundation-Models]

• Survey(IJCAI2024)(arxiv2024.02) A Comprehensive Survey of Cross-Domain Policy Transfer for Embodied Agents [arxiv link][code|official][THU]

• Survey(arxiv2024.05) A Survey on Vision-Language-Action Models for Embodied AI [arxiv link][CUHK + Huawei Noah’s Ark Lab]

• Survey(arxiv2024.05) Neural Scaling Laws for Embodied AI [arxiv link][TUM + MIT][This paper presents the first study to quantify scaling laws for Robot Foundation Models (RFMs) and the use of LLMs in robotics tasks.]

• 👍Survey(arxiv2024.04) What Foundation Models can Bring for Robot Learning in Manipulation: A Survey [arxiv link][Samsung + Beijing University of Posts and Telecommunications + Tsinghua University + Universit ̈at Hamburg, Germany]

• 👍Survey(DiffusionPolicy-Robotics)(arxiv2025.04) A Survey on Diffusion Policy for Robotic Manipulation: Taxonomy, Analysis, and Future Directions [arxiv link][code|official]

• Survey(IJCV2025)(arxiv2025.06) Vision Generalist Model: A Survey [paper link][arxiv link][Tsinghua University, China + Tencent HunyuanX, China + Beijing University of Posts and Telecommunications, China + University of Science and Technology Beijing; Jiwen Lu]

• Survey(arxiv2025.07) The Developments and Challenges towards Dexterous and Embodied Robotic Manipulation: A Survey [arxiv link][Zhejiang University]

• Survey(arxiv2025.08) Vision-Language-Action Models: Concepts, Progress, Applications and Challenges [arxiv link][Cornell University + The Hong Kong University of Science and Technology + University of the Peloponnese]

• Survey(arxiv2025.08) Large VLM-based Vision-Language-Action Models for Robotic Manipulation: A Survey [arxiv link][code|official][Harbin Institute of Technology (Shenzhen)]

• 👍FBSS(Nature-Communications-2022) Touchless interactive teaching of soft robots through flexible bimodal sensory interfaces [paper link][Beihang University; flexible bimodal smart skin (FBSS)]

• ❤👍👍ViTaM(Nature-Communications-2024) Capturing forceful interaction with deformable objects using a deep learning-powered stretchable tactile array [paper link][code|official][weixin blog][Cewu Lu][Visual-Tactile recording and tracking system for Manipulation]

• HIL-SERL(Science-Robotics-2025)(arxiv2024.10) Precise and Dexterous Robotic Manipulation via Human-in-the-Loop Reinforcement Learning [paper link][arxiv link][project link][code|official][University of California, Berkeley; Jianlan Luo + Sergey Levine]

• ❤👍CoordinatedBadminton(Science-Robotics-2025)(arxiv2025.05) Learning coordinated badminton skills for legged manipulators [paper link][arxiv link][ETH Zurich]

• ❤👍👍LEGION(Nature-Machine-Intelligence-2025) Preserving and combining knowledge in robotic lifelong reinforcement learning [paper link][project link][code|official][Technical University of Munich + Nanjing University + Sun Yat-sen University + Tsinghua University]

• ❤👍👍NeuralJacobianFields(Nature-2025)(arxiv2024.07) Controlling diverse robots by inferring Jacobian fields with deep networks [paper link][arxiv link][project link][code|official][CSAIL, MIT][The initial arxiv title of this work is Unifying 3D Representation and Control of Diverse Robots with a Single Camera]

• ❤👍PHOENIX(Nature-Communications-2025) A physics-informed and data-driven framework for robotic welding in manufacturing [paper link][code|official][Beijing University of Technology + Osaka University + Qilu University of Technology (Shandong Academy of Sciences)][Physics-informed Hybrid Optimization framework for Efficient Neural Intelligence (PHOENIX)]

• ❤👍F-TAC-Hand(Nature-Machine-Intelligence-2025)(arxiv2024.12) Embedding high-resolution touch across robotic hands enables adaptive human-like grasping [paper link][arxiv link][Peking University + Beijing Institute for General Artificial Intelligence + PKU-Wuhan Institute for Artificial Intelligenc + Queen Mary University of London; Song-Chun Zhu]

• **** [openreview link][paper link][arxiv link][project link][code|official]

This line of research may open the possibility of on-line hand-eye calibration, which is more robust and scalable then classic hand-eye calibration systems

• DREAM(ICRA2020)(arxiv2019.11) Camera-to-Robot Pose Estimation from a Single Image [paper link][arxiv link][project link][code|official][NVIDIA + CMU]

• RoMa(3DV)(arxiv2021.03) Deep Regression on Manifolds: A 3D Rotation Case Study [paper link][arxiv link][project link][code|official]

• RoboPose(CVPR2021 oral)(arxiv2021.04) Single-View Robot Pose and Joint Angle Estimation via Render & Compare [paper link][arxiv link][project link][code|official][ENS/Inria + LIGM, ENPC + CIIRC CTU]

• EasyHeC(RAL2023)(arxiv2023.05) EasyHeC: Accurate and Automatic Hand-eye Calibration via Differentiable Rendering and Space Exploration [paper link][arxiv link][project link][code|official][State Key Lab of CAD & CG, Zhejiang University + University of California, San Diego; Xiaowei Zhou + Hao Su]

• HolisticRoboPose(ECCV2024)(arxiv2024.02) Real-time Holistic Robot Pose Estimation with Unknown States [arxiv link][project link][code|official][Peking University + Shanghai Jiao Tong University]

• BoT(RSS2024 Workshop) Body Transformer: LeveragingRobot Embodiment for Policy Learning [openreview link][project link][UC Berkeley]

• Kalib(arxiv2024.08) Kalib: Markerless Hand-Eye Calibration with Keypoint Tracking [arxiv link][project link][code|official][SJTU; Cewu Lu]

• FreePose(ICRA2025)(arxiv2025.03) Foundation Feature-Driven Online End-Effector Pose Estimation: A Marker-Free and Learning-Free Approach [arxiv link][project link][code|official][Center on Frontiers of Computing Studies, School of Computer Science, Peking University; Hao Dong]

• Bi-JCR(arxiv2025.05) Bi-Manual Joint Camera Calibration and Scene Representation [arxiv link][project link][Carnegie Mellon University + Vanderbilt University]

• TACTO(RAL2022)(arxiv2020.12) TACTO: A Fast, Flexible, and Open-Source Simulator for High-Resolution Vision-Based Tactile Sensors [paper link][arxiv link][code|official][facebook][using the Tactile signal as the input]

• TactileSim(CoRL2022) Efficient Tactile Simulation with Differentiability for Robotic Manipulation [openreview link][paper link][poster link][MIT + Texas A&M University][using the Tactile signal as the input]

• Tac2Pose(IJRR2023)(arxiv2022.04) Tac2Pose: Tactile Object Pose Estimation from the First Touch [paper link][arxiv link][project link][MIT]

• See-to-Touch(ICRA2024)(arxiv2023.09) See to Touch: Learning Tactile Dexterity through Visual Incentives [[paper link](See to Touch: Learning Tactile Dexterity through Visual Incentives)][arxiv link][project link][code|official][New York University + Meta]

• RobotSynesthesia(ICRA2024)(arxiv2023.12) Robot Synesthesia: In-Hand Manipulation with Visuotactile Sensing [paper link][arxiv link][project link][code|official][UC San Diego + Tsinghua University + University of Illinois Urbana-Champaign + UC Berkeley + Dongguk University; Xiaolong Wang]

• TactGen(TRO2024) TactGen: Tactile Sensory Data Generation via Zero-Shot Sim-to-Real Transfer [paper link][Oxford University]

• 👍HATO(arxiv2024.04) Learning Visuotactile Skills with Two Multifingered Hands [arxiv link][paper link][dataset link][code|official][UC Berkeley][They repurpose two prosthetic hands with touch sensing for research use, develop a bimanual multifingered hands teleoperation system to collect visuotactile data, and learn cool policies.]

• DexSkills(IROS2024)(arxiv2024.05) DexSkills: Skill Segmentation Using Haptic Data for Learning Autonomous Long-Horizon Robotic Manipulation Tasks [arxiv link][project link][code|official][University of Edinburgh + Queen Mary University of London + Amazon ATS + University College London][Haptic Data]

• Tactile-Skin-RL(arxiv2024.07) Learning In-Hand Translation Using Tactile Skin With Shear and Normal Force Sensing [arxiv link][project link][Meta FAIR + University of Pennsylvania GRASP Lab + UC Berkeley + UW-Madison]

• 👍TacSL(arxiv2024.08) TacSL: A Library for Visuotactile Sensor Simulation and Learning [arxiv link][project link][code|official][NVIDIA Research + University of Washington; Dieter Fox][using Isaac Gym; for Visuotactile Sensor Simulation]

• TacDiffusion(arxiv2024.09) TacDiffusion: Force-domain Diffusion Policy for Precise Tactile Manipulation [arxiv link][code|official][Technical University of Munich]

• 3DTacDex(arxiv2024.09) Canonical Representation and Force-Based Pretraining of 3D Tactile for Dexterous Visuo-Tactile Policy Learning [arxiv link][project link][Peking University; Hao Dong]

• ForceMimic(arxiv2024.10) ForceMimic: Force-Centric Imitation Learning with Force-Motion Capture System for Contact-Rich Manipulation [arxiv link][project link][SJTU + Cewu Lu]

• 👍3D-ViTac(CoRL2024)(arxiv2024.10) 3D-ViTac: Learning Fine-Grained Manipulation with Visuo-Tactile Sensing [openreview link][arxiv link][project link][Columbia University + University of Illinois Urbana-Champaign + University of Washington]

• VTAO-BiManip(arxiv2025.01) VTAO-BiManip: Masked Visual-Tactile-Action Pre-training with Object Understanding for Bimanual Dexterous Manipulation [arxiv link][Zhejiang University]

• DOGlove(RSS2025)(arxiv2025.02) DOGlove: Dexterous Manipulation with a Low-Cost Open-Source Haptic Force Feedback Glove [arxiv link][project link][Tsinghua University + Shanghai Qi Zhi Institute + Shanghai AI Lab; Huazhe Xu]

• TLA(arxiv2025.03) TLA: Tactile-Language-Action Model for Contact-Rich Manipulation [arxiv link][project link][Samsung Research China - Beijing (SRC-B) + Institute of Automation, Chinese Academy of Sciences + Beijing Academy of Artificial Intelligence]

• FLEX(ICRA2025)(arxiv2025.03) FLEX: A Framework for Learning Robot-Agnostic Force-based Skills Involving Sustained Contact Object Manipulation [arxiv link][project link][code|official][Tufts University]

• Taccel(arxiv2025.04) Taccel: Scaling Up Vision-based Tactile Robotics via High-performance GPU Simulation [arxiv link][project link][code|official][Institute for AI, PKU + State Key Lab of General AI, BIGAI + AIVC Lab, UCLA]

• TacCompress(arxiv2025.05) TacCompress: A Benchmark for Multi-Point Tactile Data Compression in Dexterous Manipulation [arxiv link][PaXini Tech + SJTU]

• AdapTac-Dex(arxiv2025.05) Adaptive Visuo-Tactile Fusion with Predictive Force Attention for Dexterous Manipulation [arxiv link][project link][code|official][PKU + Ant Research; Hao Dong]

• ControlTac(arxiv2025.05) ControlTac: Force- and Position-Controlled Tactile Data Augmentation with a Single Reference Image [arxiv link][project link][University of Maryland]

• ForceVLA(arxiv2025.05) ForceVLA: Enhancing VLA Models with a Force-aware MoE for Contact-rich Manipulation [arxiv link][project link][Fudan University + Shanghai Jiao Tong University + National University of Singapore + Shanghai University + Xi’an Jiaotong University; Cewu Lu]

• CLAMP(arxiv2025.05) CLAMP: Crowdsourcing a LArge-scale in-the-wild haptic dataset with an open-source device for Multimodal robot Perception [arxiv link][project link][Cornell University + Horace Mann School]

• ManiFeel(arxiv2025.05) ManiFeel: Benchmarking and Understanding Visuotactile Manipulation Policy Learning [arxiv link][project link][Purdue University, USA]

• eFlesh(arxiv2025.06) eFlesh: Highly customizable Magnetic Touch Sensing using Cut-Cell Microstructures [arxiv link][project link][New York University]

• In-Hand-VTF(arxiv2025.06) In-Hand Object Pose Estimation via Visual-Tactile Fusion [arxiv link][Goethe Universit ̈at Frankfurt + TU Darmstadt + German Research Center for AI (DFKI)]

• Multi-Suction-Item-Picking(RSS2025, Demonstrating)(arxiv2025.06) Demonstrating Multi-Suction Item Picking at Scale via Multi-Modal Learning of Pick Success [paper link][arxiv link][Amazon Robotics]

• ViTacFormer(arxiv2025.06) ViTacFormer: Learning Cross-Modal Representation for Visuo-Tactile Dexterous Manipulation [arxiv link][University of California, Berkeley + Peking University + Sharpa; Pieter Abbeel]

• tool-use-few-shot-transfer(RAL2025)(arxiv2025.07) Few-shot transfer of tool-use skills using human demonstrations with proximity and tactile sensing [arxiv link][project link][The University of Edinburgh + Sony Group Corporation]

• FBI(arxiv2025.08) FBI: Learning Dexterous In-hand Manipulation with Dynamic Visuotactile Shortcut Policy [arxiv link][project link][Shanghai Jiao Tong University; Cewu Lu][Flow Before Imitation (FBI)]

• Bi-TacGrasping(arxiv2025.09) Learning-Based Collaborative Control for Bi-Manual Tactile-Reactive Grasping [arxiv link][Mississippi State University + Villanova University]

• Survey-Rearrangement(arxiv2020.11) Rearrangement: A Challenge for Embodied AI [arxiv link][Georgia Tech + Facebook AI Research + Simon Fraser University + Imperial College London + Princeton University + Intel Labs + UC Berkeley + Google + Allen Institute for AI + University of Washington + UC San Diego; Jia Deng + Sergey Levine + Hao Su]

• IKEA Furniture Assembly(ICRA2021) IKEA Furniture Assembly Environment for Long-Horizon Complex Manipulation Tasks [paper link][arxiv link][project link][code|official][Cognitive Learning for Vision and Robotics (CLVR), University of Southern California]

• ReorientBot(ICRA2022)(arxiv2022.02) ReorientBot: Learning Object Reorientation for Specific-Posed Placement [paper link][arxiv link][Dyson Robotics Laboratory, Imperial College London; Stephen Jame]

• IFOR(CVPR2022)(arxiv2022.02) IFOR: Iterative Flow Minimization for Robotic Object Rearrangement [paper link][arxiv link][project link][code|official][NVIDIA + Princeton University; Jia Deng + Dieter Fox]

• Factory(RSS2022)(arxiv2022.05) Factory: Fast Contact for Robotic Assembly [arxiv link][projec link][code|official][NVIDIA; Isaac Gym]

• Relational-NDF(CoRL2022)(arxiv2022.11) SE(3)-Equivariant Relational Rearrangement with Neural Descriptor Fields [openreview link][paper link][arxiv link][project link][code|official][Massachusetts Institute of Technology]

• LEGO-Net(CVPR2023)(arxiv2023.01) LEGO-Net: Learning Regular Rearrangements of Objects in Rooms [paper link][arxiv link][project link][code|official][Brown University + Stanford University]

• CabiNet(ICRA2023)(arxiv2023.04) CabiNet: Scaling Neural Collision Detection for Object Rearrangement with Procedural Scene Generation [paper link][arxiv link][project link][code|official][NVIDIA; Dieter Fox][Procedural Scene Generation]

• IndustReal(RSS2023)(arxiv2023.05) IndustReal: Transferring Contact-Rich Assembly Tasks from Simulation to Reality [arxiv link][project link][IndustRealKit link][IndustRealSim link][IndustRealLib link][University of Southern California + Stanford University + NVIDIA + University of Sydney + University of Washington; Isaac Gym][following the previous work Factory(RSS2022)]

• Selective-Obj-Rearrangement(CoRL2023) Selective Object Rearrangement in Clutter [openreview link][paper link][project link][University of Southern California]

• Bimanual-Handover-Rearrangement(ICRA2023) Efficient Bimanual Handover and Rearrangement via Symmetry-Aware Actor-Critic Learning [paper link][project link][Tsinghua University + UC San Diego + Shanghai Artificial Intelligence Lab + Shanghai Qi Zhi Institute]

• SG-Bot(ICRA2024)(arxiv2023.09) SG-Bot: Object Rearrangement via Coarse-to-Fine Robotic Imagination on Scene Graphs [paper link][arxiv link][project link][code|official][Technical University of Munich + Google]

• SeqDex(CoRL2023)(arxiv2023.09) Sequential Dexterity: Chaining Dexterous Policies for Long-Horizon Manipulation [openreview link][paper link][arxiv link][project link][code|official][Stanford University; Li Fei-Fei][Its first workspace is of Building Blocks/Assemblely task in simulation and real-world, which is a long-horizon task includes four different subtasks.][It uses the Allegro Hand to conduct their real robot experiments]

• D3Fields(CoRL2024, Oral)(arxiv2023.09) D3Fields: Dynamic 3D Descriptor Fields for Zero-Shot Generalizable Rearrangement [openreview link][arxiv link][project link][code|official][Columbia University + University of Illinois Urbana-Champaign + National University of Singapore + Boston Dynamics AI Institute + Stanford University; Jiajun Wu + Li Fei-Fei]

• ReorientDiff(ICRA2024)(arxiv2023.03) ReorientDiff: Diffusion Model based Reorientation for Object Manipulation [paper link][arxiv link][project link][Georgia Institute of Technology][Diffusion Models for language-conditioned multi-step object manipulation for precise object placement.]

• D4PAS(ICRA2024)(arxiv2023.12) Multi-level Reasoning for Robotic Assembly: From Sequence Inference to Contact Selection [paper link][arxiv link][UC Berkeley; It is a large-scale Dataset for Part Assembly Sequences (D4PAS)]

• Dream2Real(ICRA2024)(arxiv2023.12) Dream2Real: Zero-Shot 3D Object Rearrangement with Vision-Language Models [paper link][arxiv link][project link][code|official][The Robot Learning Lab at Imperial College London + The Dyson Robotics Lab at Imperial College London; Edward Johns]

• Open6DOR(IROS2024)(ICRAW2024 Oral) Open6DOR: Benchmarking Open-instruction 6-DoF Object Rearrangement and A VLM-based Approach [paper link][openreview link][project link][PKU, by the He Wang group][This is a work published in the First Vision and Language for Autonomous Driving and Robotics Workshop]

• Favor(AAAI2024) Favor: Full-Body AR-driven Virtual Object Rearrangement Guided by Instruction Text [paper link][project link][Shanghai Jiao Tong University + XREAL + South China University of Technology; Cewu Lu]

• 👍👍FMB(IJRR2024)(arxiv2024.01) FMB: A Functional Manipulation Benchmark for Generalizable Robotic Learning [arxiv link][project link][Materials and CAD Files][dataset link][code|official][University of California, Berkeley (BAIR)]

• DiffAssemble(CVPR2024)(arxiv2024.02) DiffAssemble: A Unified Graph-Diffusion Model for 2D and 3D Reassembly [paper link][arxiv link][project link][code|official][Pattern Analysis and Computer Vision (PAVIS) + Istituto Italiano di Tecnologia (IIT)][It focused on 2D and 3D reassembly tasks]

• 3DHPA(CVPR2024)(arxiv2024.02) Generative 3D Part Assembly via Part-Whole-Hierarchy Message Passing [paper link][arxiv link][code|official][Peking University + University of British Columbia + Vector Institute for AI + Canada CIFAR AI Chair]

• MultiJointAssembly(CVPR2024)(arxiv2023.03) Category-Level Multi-Part Multi-Joint 3D Shape Assembly [paper link][arxiv link][code|official][Stanford University + MIT CSAIL + NVIDIA + Tsinghua University + Peking University + National University of Singapore]

• SCANet(IROS2024, Oral)(arxiv2024.03) SCANet: Correcting LEGO Assembly Errors with Self-Correct Assembly Network [arxiv link][project link][code|official][Southeast University + Peking University; Dong Hao]

• VPI(IROS2024)(arxiv2024.03) Visual Preference Inference: An Image Sequence-Based Preference Reasoning in Tabletop Object Manipulation [paper link][arxiv link][project link][code|official][Korea University + ETRI + Neubla]

• JPDVT(CVPR2024)(arxiv2024.04) Solving Masked Jigsaw Puzzles with Diffusion Vision Transformers [paper link][arxiv link][project link][code|official][Northeastern University + Qualcomm]

• ClutterGen(CoRL2024)(arxiv2024.07) ClutterGen: A Cluttered Scene Generator for Robot Learning [openreview link][arxiv link][project link][code|official][Duke University]

• 👍AutoMate(RSS2024)(arxiv2024.07) AutoMate: Specialist and Generalist Assembly Policies over Diverse Geometries [paper link][arxiv link][project link][University of Southern California + NVIDIA Corporation + University of Washington + University of Sydney; Dieter Fox][Assembly]

• DegustaBot(arxiv2024.07) DegustaBot: Zero-Shot Visual Preference Estimation for Personalized Multi-Object Rearrangement [arxiv link][Carnegie Mellon University + Hello Robot][Multi-Object Rearrangement]

• ARCH(arxiv2024.09) ARCH: Hierarchical Hybrid Learning for Long-Horizon Contact-Rich Robotic Assembly [arxiv link][project link][Stanford University + MIT + University of Michigan + Autodesk Research]

• PACA(WACV2025)(arxiv2024.10) PACA: Perspective-Aware Cross-Attention Representation for Zero-Shot Scene Rearrangement [paper link][arxiv link][KTH Royal Institute of Technology + Graz University of Technology]

• LLM-driven-Rearrangement(arxiv2025.01) Learn from the Past: Language-conditioned Object Rearrangement with Large Language Models [arxiv link][University of York + University of Southampton]

• 👍FetchBot(CoRL2025 oral)(arxiv2025.02) FetchBot: Object Fetching in Cluttered Shelves via Zero-Shot Sim2Real [arxiv link][project link][Chinese Academy of Sciences + School of Computer Science, Peking University + Galbot + Beijing Academy of Artificial Intelligence; He Wang]

• 👍AnyPlace(CoRL2025)(arxiv2025.02) AnyPlace: Learning Generalized Object Placement for Robot Manipulation [arxiv link][project link][code|official][University of Toronto + Vector Institute + Shanghai Jiao Tong University + Wilfrid Laurier University + Acceleration Consortium + Georgia Institute of Technology]

• QBIT(arxiv2025.03) QBIT: Quality-Aware Cloud-Based Benchmarking for Robotic Insertion Tasks [arxiv link][code|official][Karlsruhe University of Applied Sciences, Karlsruhe, Germany + Karlsruhe Institute of Technology, Karlsruhe, Germany]

• Action-Prior-Alignment(TASE2025 / CoRL2025 Workshop)(arxiv2025.03) Efficient Alignment of Unconditioned Action Prior for Language-conditioned Pick and Place in Clutter [arxiv link][project link][code|official][Zhejiang University + Alibaba Cloud; Rong Xiong][Pick-and-Place in Clutter]

• SLeRP(arxiv2025.04) Slot-Level Robotic Placement via Visual Imitation from Single Human Video [arxiv link][project link][NVIDIA + Univ. of Michigan + Univ. of Washington + New York University; Dieter Fox][Slot-Level Placement]

• TwoByTwo(CVPR2025)(arxiv2025.04) Two by Two: Learning Multi-Task Pairwise Objects Assembly for Generalizable Robot Manipulation [arxiv link][project link][code|official][Shanghai Qi Zhi Institute + Northeastern University + IIIS, Tsinghua University + Shanghai Jiao Tong University + Shanghai AI Lab; Huazhe Xu]

• MasterRulesFromChaos(ICRA2025)(arxiv2025.05) Master Rules from Chaos: Learning to Reason, Plan, and Interact from Chaos for Tangram Assembly [arxiv link][project link][code|official][HKUST]

• 👍BiAssemble(ICML2025)(arxiv2025.06) BiAssemble: Learning Collaborative Affordance for Bimanual Geometric Assembly [openreview link][arxiv link][project link][code|official][rejected by ICLR2025][Peking University + PKU-Agibot Lab]

• Fabrica(CoRL2025 Oral, Best Paper Award)(arxiv2025.06) Fabrica: Dual-Arm Assembly of General Multi-Part Objects via Integrated Planning and Learning [openreview link][arxiv link][project link][MIT CSAIL + ETH Zurich + Autodesk Research + Texas A&M University]

• 👍DualAfford(ICLR2023)(arxiv2022.07) DualAfford: Learning Collaborative Visual Affordance for Dual-gripper Manipulation [openreview link][arxiv link][project link][code|official][Peking University + Stanford University + NVIDIA Research + Tencent AI Lab + BAAI; Hao Dong]

• IAGNet(ICCV2023)(arxiv2023.03) Grounding 3D Object Affordance from 2D Interactions in Images [paper link][arxiv link][project link][code|official][University of Science and Technology of China + University of Rochester + Institute of Artificial Intelligence, Hefei Comprehensive National Science Center]

• OA-Gaze(ICCV2023)(arxiv2023.07) Object-aware Gaze Target Detection [paper link][arxiv link][code|official][University of Trento, Trento, Italy + Fondazione Bruno Kessler, Trento, Italy + University of Pisa, Pisa, Italy]

• 👍RoboTAP(ICRA2024)(arxiv2023.08) RoboTAP: Tracking Arbitrary Points for Few-Shot Visual Imitation [paper link][arxiv link][project link][code|official][Google DeepMind + University College London][It is based on the TAPNet]

• AVDC(ICLR2024)(arxiv2023.10) Learning to Act from Actionless Videos through Dense Correspondences [openreview link][arxiv link][project link][code|official][National Taiwan University + MIT][This method is cited by ATM(RSS2024), and has a inferior performance then ATM]

• OOAL(CVPR2024)(arxiv2023.11) One-Shot Open Affordance Learning with Foundation Models [paper link][arxiv link][project link][code|official][University of Edinburgh + Google Research + Stability AI]

• 👍Robo-ABC(ECCV2024)(arxiv2024.01) Robo-ABC: Affordance Generalization Beyond Categories via Semantic Correspondence for Robot Manipulation [arxiv link][project link][code|official][Shanghai Qi Zhi Institute + THU + SJTU]

• 👍ATM(RSS2024)(arxiv2024.01) Any-point Trajectory Modeling for Policy Learning [paper link][arxiv link][project link][code|official][UC Berkeley + IIIS, Tsinghua University + Stanford University + Shanghai Artificial Intelligence Laboratory + Shanghai Qi Zhi Institute + CUHK][The method is evaluated on a challenging simulation benchmark (LIBERO) comprised of 130 language-conditioned manipulation tasks, and on 5 tasks in a real-world UR5 Kitchen environment.]

• 👍MOKA(RSS2024)(arxiv2024.03) MOKA: Open-World Robotic Manipulation through Mark-based Visual Prompting [paper link][arxiv link][project link][code|official][Berkeley AI Research, UC Berkeley; Pieter Abbeel + Sergey Levine]

• Bi-KVIL(ICRA2024)(arxiv2024.03) Bi-KVIL: Keypoints-based Visual Imitation Learning of Bimanual Manipulation Tasks [paper link][arxiv link][project link][code|official][Karlsruhe Institute of Technology][The proposed Bi-KVIL jointly extracts so-called Hybrid Master-Slave Relationships (HMSR) among objects and hands, bimanual coordination strategies, and sub-symbolic task representations]

• KAT(RSS2024)(arxiv2024.03) Keypoint Action Tokens Enable In-Context Imitation Learning in Robotics [paper link][arxiv link][project link][code|official][The Robot Learning Lab, Imperial College London; Edward Johns]

• PreAfford(IROS2024)(arxiv2024.04) PreAfford: Universal Affordance-Based Pre-Grasping for Diverse Objects and Environments [paper link][arxiv link][project link][code|official][THU + PKU]

• IMOP(RSS2024)(arxiv2024.05) One-Shot Imitation Learning with Invariance Matching for Robotic Manipulation [paper link][arxiv link][project link][code|official][Rutgers University, Invariance-Matching One-shot Policy Learning (IMOP)][Render&Diffuse][only tested on the dataset RLBench, and obtained inferior results than 3D Diffuser Actor][Learning from action labels free human videos]

• 👍Track2Act(ECCV2024)(arxiv2024.05) Track2Act: Predicting Point Tracks from Internet Videos Enables Generalizable Robot Manipulation [paper link][arxiv link][project link][code|official][Carnegie Mellon University + University of Washington + Meta][The first author is Homanga Bharadhwaj who has given a position paper in ICML2024 named Position: Scaling Simulation is Neither Necessary Nor Sufficient for In-the-Wild Robot Manipulation]

• 👍RoboPoint(CoRL2024)(arxiv2024.06) RoboPoint: A Vision-Language Model for Spatial Affordance Prediction for Robotics [openreview link][arxiv link][project link][code|official][University of Washington + NVIDIA; Dieter Fox][ROBOPOINT is a VLM that predicts image keypoint affordances given language instructions.]

• 👍RAM(CoRL2024, Oral)(arxiv2024.07) RAM: Retrieval-Based Affordance Transfer for Generalizable Zero-Shot Robotic Manipulation [openreview link][arxiv link][project link][code|official][University of Southern California + Peking University + Stanford University; He Wang]

• MIFAG(AAAI2025)(arxiv2024.08) Learning 2D Invariant Affordance Knowledge for 3D Affordance Grounding [paper link][arxiv link][project link][code|official][University of Science and Technology of China + Shanghai AI Laboratory + Northwestern Polytechnical University + TeleAI, China Telecom Corp Ltd][Multi-Image Guided Invariant-Feature-Aware 3D Affordance Grounding (MIFAG)]

• Aff-Grasp(ICCV2025)(arxiv2024.08) Learning Precise Affordances from Egocentric Videos for Robotic Manipulation [arxiv link][project link][code|official][University of Edinbugh + Huawei Noah’s Ark Lab]

• FlowMatchingPolicy(CoRL2024 Workshop)(arxiv2024.09) Affordance-based Robot Manipulation with Flow Matching [openreview link][arxiv link][project link][Honda Research Institute EU]

• 👍👍ReKep(CoRL2024)(arxiv2024.09) ReKep: Spatio-Temporal Reasoning of Relational Keypoint Constraints for Robotic Manipulation [openreview link][arxiv link][project link][code|official][Stanford University + Columbia University; Li Fei-Fei]

• 👍affordance-policy(arxiv2024.10) Affordance-Centric Policy Learning: Sample Efficient and Generalisable Robot Policy Learning using Affordance-Centric Task Frames [arxiv link][project link][QUT Centre for Robotics + University of Adelaide]

• RT-Affordance(CoRL2024 Workshop)(arxiv2024.11) RT-Affordance: Affordances are Versatile Intermediate Representations for Robot Manipulation [openreview link][arxiv link][project link][Google DeepMind + The University of Austin at Texas; Yuke Zhu]

• AffordDP(CVPR2025)(arxiv2024.12) AffordDP: Generalizable Diffusion Policy with Transferable Affordance [paper link][arxiv link][project link][ShanghaiTech University]

• P3PO(arxiv2024.12) P3-PO: Prescriptive Point Priors for Visuo-Spatial Generalization of Robot Policies [arxiv link][project link][code|official][University of Maryland, College Park + New York University]

• 👍DenseMatcher(ICLR2025, Spotlight)(arxiv2024.12) DenseMatcher: Learning 3D Semantic Correspondence for Category-Level Manipulation from a Single Demo [openreview link][arxiv link][project link][code|official][IIIS, Tsinghua University + Tepan Inc. + Shanghai Qi Zhi Institute + UC Berkeley + Stanford University + Shanghai AI Lab + Shanghai Jiao Tong University + Huazhe Xu]

• 👍👍GazeTD(arxiv2025.01) Gaze-based Task Decomposition for Robot Manipulation in Imitation Learning [arxiv link][code|official][The University of Tokyo]

• 👍SKIL(RSS2025)(arxiv2025.01) SKIL: Semantic Keypoint Imitation Learning for Generalizable Data-efficient Manipulation [paper link][arxiv link][project link][Tsinghua University + Shanghai AI Laboratory + Shanghai Qi Zhi Institute + Yang Gao]

• 👍Point-Policy(CoRL2025)(arxiv2025.02) Point Policy: Unifying Observations and Actions with Key Points for Robot Manipulation [openreview link][paper link][arxiv link][project link][code|official][New York University]

• 👍IKER-Robot(ICRA2025)(arxiv2025.02) A Real-to-Sim-to-Real Approach to Robotic Manipulation with VLM-Generated Iterative Keypoint Rewards [arxiv link][project link][code|official][University of Illinois at Urbana-Champaign + Stanford University + Amazon + Columbia University; Li Fei-Fei]

• KUDA(ICRA2025)(arxiv2025.03) KUDA: Keypoints to Unify Dynamics Learning and Visual Prompting for Open-Vocabulary Robotic Manipulation [arxiv link][project link][code|official][Tsinghua University + University of Illinois Urbana-Champaign + Columbia University]

• AffordDexGrasp(arxiv2025.03) AffordDexGrasp: Open-set Language-guided Dexterous Grasp with Generalizable-Instructive Affordance [arxiv link][project link][Sun Yat-sen University]

• GarmentPile(CVPR2025)(arxiv2025.03) GarmentPile: Point-Level Visual Affordance Guided Retrieval and Adaptation for Cluttered Garments Manipulation [arxiv link][project link][code|official][PKU; Hao Dong]

• PointArena(arxiv2025.05) PointArena: Probing Multimodal Grounding Through Language-Guided Pointing [arxiv link][project link][code|official][University of Washington + Allen Institute for Artificial Intelligence + Anderson Collegiate Vocational Institute; Dieter Fox]

• 👍UAD(ICRA2025, Best Paper Finalist)(arxiv2025.06) UAD: Unsupervised Affordance Distillation for Generalization in Robotic Manipulation [arxiv link][project link][code|official][Stanford University; Jiajun Wu + Li Fei-Fei]

• HMD-Ego(arxiv2025.06) Where Do We Look When We Teach? Analyzing Human Gaze Behavior Across Demonstration Devices in Robot Imitation Learning [arxiv link][Toyota Motor Corporation + Nara Institute of Science and Technology]

• RAGNet(ICCV2025)(arxiv2025.07) RAGNet: Large-scale Reasoning-based Affordance Segmentation Benchmark towards General Grasping [arxiv link][project link][code|official][The Chinese University of Hong Kong + Institute of Computing Technology, Chinese Academy of Sciences + Dexmal + Mohamed bin Zayed University of Artificial Intelligence + SKL-IOTSC, CIS, University of Macau]

• GIAVA(arxiv2025.07) Look, Focus, Act: Efficient and Robust Robot Learning via Human Gaze and Foveated Vision Transformers [arxiv link][project link][code|official][University of California, Berkeley + Tongji University + University of California, Davis][GIAVA (Gaze Integrated Active-Vision ALOHA)]

• AffordDex(arxiv2025.08) Towards Affordance-Aware Robotic Dexterous Grasping with Human-like Priors [arxiv link][project link][code|official][Wuhan University + DAMO Academy, Alibaba Group + Hupan Lab + Zhejiang University + Tsinghua University]

• O3Afford(CoRL2025)(arxiv2025.09) O3Afford: One-Shot 3D Object-to-Object Affordance Grounding for Generalizable Robotic Manipulation [arxiv link][project link][code|official][University of Virginia]

• DexPilot(ICRA2020)(arxiv2019.10) DexPilot: Vision Based Teleoperation of Dexterous Robotic Hand-Arm System [paper link][arxiv link][project link][NVIDIA + CMU; Dieter Fox]

• (RAL2022)(arxiv2022.12) From One Hand to Multiple Hands: Imitation Learning for Dexterous Manipulation from Single-Camera Teleoperation [paper link][arxiv link][project link][code|official][UC San Diego; Hao Su + Xiaolong Wang]

• 👍👍ACT/ALOHA(RSS2023)(arxiv2023.04) Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware [paper link][arxiv link][project link][code|official][Stanford University + UC Berkeley + Meta][It adopts a CVAE scheme with transformer backbones and ResNet image encoders to model the variability of human data]

• 👍AnyTeleop(RSS2023)(arxiv2023.07) AnyTeleop: A General Vision-Based Dexterous Robot Arm-Hand Teleoperation System [paper link][arxiv link][project link][code|official][UC San Diego + NVIDIA; related to Xiaolong Wang group][This work can be used for dex-retargeting]

• Kinematic-Motion-Retargeting(arxiv2024.02) Kinematic Motion Retargeting for Contact-Rich Anthropomorphic Manipulations [arxiv link][Carnegie Mellon University + Boston Dynamics AI Institute + FAIR at Meta][Hand motion capture and retargeting]

• 👍DexCap(RSS2024)(arxiv2024.03) DexCap: Scalable and Portable Mocap Data Collection System for Dexterous Manipulation [arxiv link][project link][code|official][Stanford; Li Fei-Fei][It is a portable hand motion capture system, alongside DexIL, a novel imitation algorithm for training dexterous robot skills directly from human hand mocap data.][It showcases the system's capability to effectively learn from in-the-wild mocap data, paving the way for future data collection methods for dexterous manipulation.]

• EVE(UIST2024)(arxiv2024.04) EVE: Enabling Anyone to Train Robots using Augmented Reality [paper link][arxiv link][University of Washington + NVIDIA + Dieter Fox]

• ALOHA2(arxiv2024.05) ALOHA 2: An Enhanced Low-Cost Hardware for Bimanual Teleoperation [arxiv link][project link][code|official][Google DeepMind + Stanford University + Hoku Lab]

• Open-TeleVision(CoRL2024)(arxiv2024.07) Open-TeleVision: Teleoperation with Immersive Active Visual Feedback [openreview link][arxiv link][project link][code|official][UC San Diego + MIT; Xiaolong Wang]

• Bunny-VisionPro(arxiv2024.07) Bunny-VisionPro: Real-Time Bimanual Dexterous Teleoperation for Imitation Learning [arxiv link][project link][code|official][The University of Hong Kong + University of California, San Diego; Xiaolong Wang]

• ACE/ACETeleop(CoRL2024)(arxiv2024.08) ACE: A Cross-platform Visual-Exoskeletons for Low-Cost Dexterous Teleoperation [openreview link][arxiv link][project link][code|official][UC San Diego; Xiaolong Wang][Cross-Platform Teleoperation + Autonomous Skills]

• Haptic-ACT(arxiv2024.09) Haptic-ACT: Bridging Human Intuition with Compliant Robotic Manipulation via Immersive VR [arxiv link][project link][Robot Intelligence Lab, Imperial College London + Extend Robotics]

• ARCap(arxiv2024.10) ARCap: Collecting High-quality Human Demonstrations for Robot Learning with Augmented Reality Feedback [arxiv link][project link][code|official][Stanford University; Li Fei-Fei][This is a portable data collection system that provides visual feedback through augmented reality (AR) and haptic warnings to guide users in collecting high-quality demonstrations.]

• ARCADE(IROS2024)(arxiv2024.10) ARCADE: Scalable Demonstration Collection and Generation via Augmented Reality for Imitation Learning [arxiv link][project link][code|official][weixin blog][University of North Carolina at Chapel Hill][using Augmented Reality for collecting demonstrations]

• DexHub&DART(arxiv2024.11) DexHub and DART: Towards Internet Scale Robot Data Collection [arxiv link][project link][MIT][Dexterous Augmented Reality Teleoperation based on the Apple Vision Pro]

• LookAround(arxiv2024.11) Learning to Look Around: Enhancing Teleoperation and Learning with a Human-like Actuated Neck [arxiv link][MIT][Tiis idea is similar to AV-ALOHA]

• ARMADA(arxiv2024.12) ARMADA: Augmented Reality for Robot Manipulation and Robot-Free Data Acquisition [arxiv link][project link][Apple]

• TelePreview(arxiv2024.12) TelePreview: A User-Friendly Teleoperation System with Virtual Arm Assistance for Enhanced Effectiveness [arxiv link][project link][National University of Singapore]

• TeleOpBench(arxiv2025.05) TeleOpBench: A Simulator-Centric Benchmark for Dual-Arm Dexterous Teleoperation [arxiv link][project link][Shanghai Artificial Intelligence Laboratory + Zhejiang University + The Chinese University of Hong Kong + The Hong Kong University of Science and Technology (Guangzhou) + The University of Hong Kong + Feeling AI; Jiangmiao Pang]

• XRoboToolkit(arxiv2025.08) XRoboToolkit: A Cross-Platform Framework for Robot Teleoperation [arxiv link][project link][code|official][ByteDance + Georgia Institute of Technology + George Mason University]

• 👍❤DiffusionPolicy(RSS2023)(IJRR2024)(arxiv2023.03) Diffusion Policy: Visuomotor Policy Learning via Action Diffusion [paper link][arxiv link][project link][code|official][Columbia University + Toyota Research Institute + MIT][Did not conduct experiments on benchmarks RLBench and CALVIN]

• ❤ChainedDiffuser(CoRL2023) ChainedDiffuser: Unifying Trajectory Diffusion and Keypose Prediction for Robotic Manipulation [openreview link][paper link][project link][code|official][CMU, using the Diffusion; the first authors Zhou Xian and Nikolaos Gkanatsios][It proposed to replace motion planners, commonly used for keypose to keypose linking, with a trajectory diffusion model that conditions on the 3D scene feature cloud and the predicted target 3D keypose to denoise a trajectory from the current to the target keypose.]

• 👍Diffusion-EDFs(CVPR2024 Highlight)(arxiv2023.09) Diffusion-EDFs: Bi-equivariant Denoising Generative Modeling on SE(3) for Visual Robotic Manipulation [paper link][arxiv link][project link][code|official][Yonsei University + University of California, Berkeley + Samsung Research + MIT][It compared to the SE(3)-DiffusionFields(ICRA2023)]

• EquivAct(ICRA2024)(arxiv2023.10) EquivAct: SIM(3)-Equivariant Visuomotor Policies beyond Rigid Object Manipulation [arxiv link][project link][Stanford University + Princeton University]

• SkillDiffuser(CVPR2024)(arxiv2023.12) SkillDiffuser: Interpretable Hierarchical Planning via Skill Abstractions in Diffusion-Based Task Execution [paper link][arxiv link][project link][code|official][The University of Hong Kong + UC Berkeley + Shanghai AI Laboratory; Ping Luo]

• Diff-Control(IROS2024) Diff-Control: A Stateful Diffusion-based Policy for Imitation Learning [pdf link][project link][code|official][Interactive Robotics Lab, Arizona State University + Kyushu Institute of Technology]

• dmd_diffusion(RSS2024)(arxiv2024.02) Diffusion Meets DAgger: Supercharging Eye-in-hand Imitation Learning [arxiv link][project link][code|official][University of Illinois at Urbana-Champaign]

• 👍👍3D Diffuser Actor(CoRL2024)(arxiv2024.02) 3D Diffuser Actor: Policy Diffusion with 3D Scene Representations [openreview link][arxiv link][project link][code|official][CMU, using the Diffusion; the first authors Tsung-Wei Ke and Nikolaos Gkanatsios][This work is largely based on their previous work Actor3D and ChainedDiffuser, and also closely related with methods PerAct, DiffusionPolicy, RVT and GNFactor][It used rotary positional embeddings proposed by RoFormer to bulid the 3D Relative Position Denoising Transformer module.][Comparing to ChainedDiffuser, It instead predicts the next 3D keypose for the robot’s end-effector alongside the linking trajectory, which is a much harder task than linking two given keyposes.][The previous version 3D Diffuser Actor is rejected by ICLR2024 for being similar to Actor3D.]

• ❤HDP(CVPR2024)(arxiv2024.03) Hierarchical Diffusion Policy for Kinematics-Aware Multi-Task Robotic Manipulation [arxiv link][project link][code|official][Dyson Robot Learning Lab][It uses PerAct as the high-level agent]

• ❤DP3(RSS2024)(arxiv2024.03) 3D Diffusion Policy: Generalizable Visuomotor Policy Learning via Simple 3D Representations [arxiv link][project link][code|official][Shanghai Qizhi + SJTU + THU + Shanghai AI Lab][This work is also published on IEEE 2024 ICRA Workshop 3D Manipulation.][Did not conduct experiments on benchmarks RLBench and CALVIN]

• DNAct(arxiv2024.03) DNAct: Diffusion Guided Multi-Task 3D Policy Learning [arxiv link][project link][UC San Diego; a work by the Xiaolong Wang group][It leverages neural rendering to distill 2D semantic features from foundation models such as Stable Diffusion to a 3D space, which provides a comprehensive semantic understanding regarding the scene.]

• ❤ConsistencyPolicy(RSS2024)(arxiv2024.05) Consistency Policy: Accelerated Visuomotor Policies via Consistency Distillation [arxiv link][project link][code|official][Stanford University + Princeton University; Consistency Policy accelerates Diffusion Policy for real time inference on compute constrained robotics platforms.`]

• R&D(RSS2024)(arxiv2024.05) Render and Diffuse: Aligning Image and Action Spaces for Diffusion-based Behaviour Cloning [arxiv link][project link][Dyson Robot Learning Lab + Imperial College London][It compared to methods ACT and Diffusion Policy on RLBench; It did not consider adding the 3D information into inputs.]

• 👍ManiCM(arxiv2024.06) ManiCM: Real-time 3D Diffusion Policy via Consistency Model for Robotic Manipulation [arxiv link][project link][code|official][THU-SZ + Shanghai AI Lab + CMU][It is based on 3D Diffusion Policy and is much better, where DP3 is accelerated via consistency model.][It did not conduct experiments on benchmarks RLBench and CALVIN]

• Streaming-DP(arxiv2024.06) Streaming Diffusion Policy: Fast Policy Synthesis with Variable Noise Diffusion Models [arxiv link][project link][code|official][Norwegian University of Science and Technology + Harvard University]

• EquiDiff(CoRL2024, Outstanding Paper Award Finalist)(arxiv2024.07) Equivariant Diffusion Policy [openreview link][arxiv link][project link][code|official][Northeastern University + Boston Dynamics AI Institute]

• MDT-Policy(RSS2024)(arxiv2024.07) Multimodal Diffusion Transformer: Learning Versatile Behavior from Multimodal Goals [arxiv link][project link][code|official][Intuitive Robots Lab (IRL), Karlsruhe Institute of Technology][It tested on benchmarks CALVIN and LIBERO.]

• 👍EquiBot(CoRL2024)(arxiv2024.07) EquiBot: SIM(3)-Equivariant Diffusion Policy for Generalizable and Data Efficient Learning [openreview link][arxiv link][project link][code|official][Stanford University][ This work is largely based on their previous work (ICRA2024)(arxiv2023.10) EquivAct: SIM(3)-Equivariant Visuomotor Policies beyond Rigid Object Manipulation.][During the human demonstration processing stage, it used Grounded Segment Anything Model with DEVA (ICCV2023) as the object detection and tracking model and HaMeR (CVPR2024) as the hand detection model.]

• BiD-Diffusion(arxiv2024.08) Bidirectional Decoding: Improving Action Chunking via Closed-Loop Resampling [arxiv link][project link][code|official][Stanford University; Chelsea Finn]

• Scalable-DP(arxiv2024.09) Scalable Diffusion Policy: Scale Up Diffusion Policy via Transformers for Visuomotor Learning [arxiv link][project link][Midea Group + East China Normal University + Standford University, + Shanghai University]

• 👍DDPO(arxiv2024.09) Diffusion Policy Policy Optimization [arxiv link][project link][code|official][Princeton University + Massachusetts Institute of Technology + Toyota Research Institute + Carnegie Mellon University][It is an algorithmic framework and set of best practices for fine-tuning diffusion-based policies in continuous control and robot learning tasks. DPPO shows marked improvements over diffusion and non-diffusion baselines alike, across a variety of tasks and sim-to-real transfer.]

• GenDP(CoRL2024, Oral)(arxiv2024.10) GenDP: 3D Semantic Fields for Category-Level Generalizable Diffusion Policy [openreview link][arxiv link][project link][code|official][Columbia University + University of Illinois Urbana-Champaign + Boston Dynamics AI Institute][This work is based on Diffusion Policy, robomimic and D3Fields]

• Shortcut-models(arxiv2024.10) One Step Diffusion via Shortcut Models [arxiv link][project link][code|official][UC Berkeley; Sergey Levine + Pieter Abeel]

• 👍DiT-Block-Policy(arxiv2024.10) The Ingredients for Robotic Diffusion Transformers [arxiv link][project link][code|official][Carnegie Mellon University + UC Berkeley; Sergey Levine]

• ET-SEED(ICLR2025)(arxiv2024.11) ET-SEED: Efficient Trajectory-Level SE(3) Equivariant Diffusion Policy [openreview link][arxiv link][project link][code|official][Peking University + National University of Singapore; Hao Dong]

• CARP(arxiv2024.12) CARP: Visuomotor Policy Learning via Coarse-to-Fine Autoregressive Prediction [arxiv link][project link][Westlake University + Zhejiang University + Beijing Advanced Innovation Center for Future Blockchain and Privacy Computing]

• GlideManip(arxiv2024.12) Planning-Guided Diffusion Policy Learning for Generalizable Contact-Rich Bimanual Manipulation [arxiv link][project link][Boston Dynamics AI Institute + UC San Diego + Cornell University]

• MPD(arxiv2024.12) Motion Planning Diffusion: Learning and Adapting Robot Motion Planning with Diffusion Models [arxiv link][project link][Technical University of Darmstadt, Germany + Poznan University of Technology, Poland + IDEAS NCBR, Warsaw, Poland + ][SUBMITTED TO IEEE TRANSACTIONS ON ROBOTICS]

• IMLE-Policy(arxiv2025.02) IMLE Policy: Fast and Sample Efficient Visuomotor Policy Learning via Implicit Maximum Likelihood Estimation [arxiv link][project link][QUT Centre for Robotics + Sydekick Robotics]

• S2-Diffusion(arxiv2025.02) S2-Diffusion: Generalizing from Instance-level to Category-level Skills in Robot Manipulation [arxiv link][KTH Royal Institute of Technology, Sweden + INCAR Robotics AB, Sweden]

• DTP(arxiv2025.02) Diffusion Trajectory-guided Policy for Long-horizon Robot Manipulation [arxiv link][BeiHang University + Beijing Innovation Center of Humanoid Robotics + KTH Royal Institute of Technology, Sweden][Diffusion Trajectory-guided Policy (DTP)]

• DRIFT(RSS2025)(arxiv2025.02) Dynamic Rank Adjustment in Diffusion Policies for Efficient and Flexible Training [arxiv link][Yale University + University of Pennsylvania]

• KStarDiffuser(CVPR2025)(arxiv2025.03) Spatial-Temporal Graph Diffusion Policy with Kinematic Modeling for Bimanual Robotic Manipulation [paper link][arxiv link][Harbin Institute of Technology (Shenzhen) + Greate Bay University + Huawei Noah’s Ark Lab + Shandong Computer Science Center]

• 👍RDP(RSS2025)(arxiv2025.03) Reactive Diffusion Policy: Slow-Fast Visual-Tactile Policy Learning for Contact-Rich Manipulation [paper link][arxiv link][project link][code|official][Shanghai Jiao Tong University + Tsinghua University, IIIS + Shanghai Qi Zhi Institute + Shanghai AI Lab 5Shanghai Innovation Institute + Huazhe Xu + Cewu Lu]

• Falcon(ICML2025)(arxiv2025.03) Fast Visuomotor Policies via Partial Denoising [paper link][arxiv link][Peking University + BIGAI]

• 👍DensePolicy/DspNet(arxiv2025.03) Dense Policy: Bidirectional Autoregressive Learning of Actions [arxiv link][project link][code|official-1][code|official-2][Shanghai Jiao Tong University + Xidian University + Shanghai Innovation Institute; Haoshu Fang + Yong-Lu Li + Cewu Lu + Lixin Yang]

• WaveletPolicy(arxiv2025.04) Wavelet Policy: Imitation Policy Learning in Frequency Domain with Wavelet Transforms [arxiv link][Zhejiang University + Tsinghua Universit]

• 3D-EVP(arxiv2025.05) 3D Equivariant Visuomotor Policy Learning via Spherical Projection [arxiv link][project link][Northeastern University]

• ADCS(arxiv2025.05) Adaptive Diffusion Constrained Sampling for Bimanual Robot Manipulation [arxiv link][project link][TU Darmstadt + Hessian.AI + Robotics Institute Germany]

• RoboTransfer(arxiv2025.05) RoboTransfer: Geometry-Consistent Video Diffusion for Robotic Visual Policy Transfer [arxiv link][project link][Horizon Robotics + GigaAI + CASIA]

• EquAct(arxiv2025.05) EquAct: An SE(3)-Equivariant Multi-Task Transformer for Open-Loop Robotic Manipulation [arxiv link][Northeastern University]

• CanonicalPolicy(arxiv2025.05) Canonical Policy: Learning Canonical 3D Representation for Equivariant Policy [arxiv link][project link][Purdue University]

• CDM(ICRA2025)(arxiv2025.05) Cascaded Diffusion Models for Neural Motion Planning [arxiv link][Carnegie Mellon University + University of Washington + Hello Robot]

• DemoSpeedup(arxiv2025.06) DemoSpeedup: Accelerating Visuomotor Policies via Entropy-Guided Demonstration Acceleration [arxiv link][project link][Shanghai Qi Zhi Institute + Tsinghua Embodied AI Lab @ IIIS, Tsinghua University + Shanghai AI Lab + University of Electronic Science and Technology of China; Huazhe Xu]

• LiPo(arxiv2025.06) LiPo: A Lightweight Post-optimization Framework for Smoothing Action Chunks Generated by Learned Policies [arxiv link][project link][code|official][Kwangwoon University]

• 👍RTC(arxiv2025.06) Real-Time Execution of Action Chunking Flow Policies [arxiv link][project link][Physical Intelligence; Sergey Levine][real-time chunking (RTC)]

• FreqPolicy(arxiv2025.06) FreqPolicy: Efficient Flow-based Visuomotor Policy via Frequency Consistency [arxiv link][Beijing Innovation Center of Humanoid Robotics + NLPR, MAIS, Institute of Automation of Chinese Academy of Sciences]

• RIP(IROS2025)(arxiv2025.06) Robust Instant Policy: Leveraging Student's t-Regression Model for Robust In-context Imitation Learning of Robot Manipulation [arxiv link][project link][National Institute of Advanced Industrial Science and Technology (AIST), Japan]

• CDP(arxiv2025.06) CDP: Towards Robust Autoregressive Visuomotor Policy Learning via Causal Diffusion [arxiv link][Sun Yat-sen University + CUHK(SZ)]

• DP4(arxiv2025.07) Spatial-Temporal Aware Visuomotor Diffusion Policy Learning [arxiv link][project link][Fudan University + Shanghai Innovation Institute + Nanyang Technological University + NeuHelium Co., Ltd][Spatial-Temporal Aware Visuomotor Diffusion Policy Learning (4D Diffusion Policy)]

• RTI-DP(IROS2025)(arxiv2025.08) Real-Time Iteration Scheme for Diffusion Policy [arxiv link][project link][code|official][KTH Royal Institute of Technology + University of Copenhagen]

• CP-Gen(CoRL2025)(arxiv2025.08) Constraint-Preserving Data Generation for Visuomotor Policy Learning [arxiv link][project link][code|official][Stanford University + University of Texas at Austin + Princeton University; Jimmy Wu; Yuke Zhu]

• LightDP(ICCV2025)(arxiv2025.08) On-Device Diffusion Transformer Policy for Efficient Robot Manipulation [arxiv link][project link][The University of Hong Kong + Westlake University + University of Newcastle + UBTech Robotics Corp.]

• 👍RT-1(RSS2023)(arxiv2022.12) RT-1: Robotics Transformer for Real-World Control at Scale [paper link][arxiv link][project link][released datasets][code|official][by Google DeepMind]

• 👍RT-2(CoRL2023)(arxiv2023.07) RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control [openreview link][paper link][arxiv link][project link][code|not official][by Google DeepMind; based on RT-1; it is trained on top of PaLM-E (12B); it is also trained on top of PaLI-X (55B); it plans to use more powerful VLMs, such as LLaVA (Large Language and Vision Assistant) and LLaVA-1.5]

• RT-H(RSS2024)(arxiv2024.03) RT-H: Action Hierarchies using Language [paper link][arxiv link][project link][blog|weixin][by Google DeepMind and Stanford University][Its insight is to teach the robot the language of actions]

• 👍Octo(RSS2024)(arxiv2024.05) Octo: An Open-Source Generalist Robot Policy [paper link][arxiv link][project link][code|official][UC Berkeley + Stanford + CMU + Google DeepMind][based on RT-1-X and RT-2-X; the low-level action policy is based on Diffusion Policy]

• 👍OpenVLA(CoRL2024)(arxiv2024.06) OpenVLA: An Open-Source Vision-Language-Action Model [paper link][arxiv link][project link][code|official][SimplerEnv-OpenVLA (not officially)][Stanford University + UC Berkeley + Toyota Research Institute + Google DeepMind + Physical Intelligence + MIT][It has better performance than RT-1/2/H/X and Octo]

• TinyVLA(RAL2025)(arxiv2024.09) TinyVLA: Towards Fast, Data-Efficient Vision-Language-Action Models for Robotic Manipulation [paper link][arxiv link][project link][Midea Group + East China Normal University + Shanghai University + Syracuse University + Beijing Innovation Center of Humanoid Robotics]

• GR-2(arxiv2024.10) GR-2: A Generative Video-Language-Action Model with Web-Scale Knowledge for Robot Manipulation [arxiv link][project link][Robotics Research Team, ByteDance Research]

• 👍π0(RSS2025)(arxiv2024.10) π0: A Vision-Language-Action Flow Model for General Robot Control [paper link][pdf link][arxiv link][project link][Physical Intelligence (π); Chelsea Finn + Sergey Levine]

• VLA-Diffu-Switch(arxiv2024.10) Vision-Language-Action Model and Diffusion Policy Switching Enables Dexterous Control of an Anthropomorphic Hand [arxiv link][project link][EPFL]

• 👍RDT-1B(ICLR2025)(arxiv2024.10) RDT-1B: a Diffusion Foundation Model for Bimanual Manipulation [openreview link][arxiv link][project link][code|official][Tsinghua University; Jun Zhu]

• 👍CogACT(arxiv2024.11) CogACT: A Foundational Vision-Language-Action Model for Synergizing Cognition and Action in Robotic Manipulation [arxiv link][project link][code|official][Microsoft Research Asia + Tsinghua University + USTC + Institute of Microelectronics, CAS]

• CoA-VLA(arxiv2024.12) Improving Vision-Language-Action Models via Chain-of-Affordance [arxiv link][project link][Midea Group + Shanghai University + East China Normal University]

• Shake-VLA(HRI2025)(arxiv2025.01) Shake-VLA: Vision-Language-Action Model-Based System for Bimanual Robotic Manipulations and Liquid Mixing [arxiv link][Skoltech, Moscow, Russia]

• SpatialVLA(arxiv2025.01) SpatialVLA: Exploring Spatial Representations for Visual-Language-Action Model [arxiv link][project link][code|official][Shanghai AI Laboratory + ShanghaiTech + TeleAI]

• 👍FAST(RSS2025)(arxiv2025.01) FAST: Efficient Action Tokenization for Vision-Language-Action Models [paper link][arxiv link][project link][code|official][Physical Intelligence + UC Berkeley + Stanford + Chelsea Finn + Sergey Levine]

• DexVLA(arxiv2025.02) DexVLA: Vision-Language Model with Plug-In Diffusion Expert for General Robot Control [arxiv link][project link][code|official][Midea Group + East China Normal University + Shanghai University]

• ChatVLA(arxiv2025.02) ChatVLA: Unified Multimodal Understanding and Robot Control with Vision-Language-Action Model [arxiv link][project link][Midea Group + East China Normal University + Shanghai University + Beijing Innovation Center of Humanoid Robotics + Tsinghua University]

• 👍OpenVLA-OFT(RSS2025)(arxiv2025.02) Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success [paper link][arxiv link][project link][code|official][Stanford University; Chelsea Finn]

• VLAS(ICLR2025)(arxiv2025.02) VLAS: Vision-Language-Action Model With Speech Instructions For Customized Robot Manipulation [openreview link][arxiv link][code|official][Westlake University + Zhejiang University + Xi’an Jiaotong University]

• HybridVLA(arxiv2025.03) HybridVLA: Collaborative Diffusion and Autoregression in a Unified Vision-Language-Action Model [arxiv link][project link][code|official][Peking University + Beijing Academy of Artificial Intelligence (BAAI) + CUHK]

• 👍π0.5(arxiv2025.04) π0.5: a Vision-Language-Action Model with Open-World Generalization [arxiv link][project link][Physical Intelligence; Sergey Levine]

• AGNOSTOS(arxiv2025.05) Exploring the Limits of Vision-Language-Action Manipulations in Cross-task Generalization [arxiv link][project link][code|official][HKUST(GZ) + HKU + SYSU + HKUST]

• VLA-RL(arxiv2025.05) VLA-RL: Towards Masterful and General Robotic Manipulation with Scalable Reinforcement Learning [arxiv link][project link][code|official][Tsinghua University + Nanyang Technological University]

• SimpleVLA-RL(year2025.05) Online RL with Simple Reward Enables Training VLA Models with Only One Trajectory [code|official][THU]

• 👍π0.5 + KI(arxiv2025.05) Knowledge Insulating Vision-Language-Action Models: Train Fast, Run Fast, Generalize Better [arxiv link][project link][Physical Intelligence; Sergey Levine]

• FlashVLA(arxiv2025.05) Think Twice, Act Once: Token-Aware Compression and Action Reuse for Efficient Inference in Vision-Language-Action Models [arxiv link][Fudan University + Shanghai AI Laboratory + The Chinese University of Hong Kong + Zhangjiang Laboratory]

• ChatVLA-2(arxiv2025.05) ChatVLA-2: Vision-Language-Action Model with Open-World Embodied Reasoning from Pretrained Knowledge [arxiv link][project link][Midea Group + East China Normal University]

• OneTwoVLA(arxiv2025.05) OneTwoVLA: A Unified Vision-Language-Action Model with Adaptive Reasoning [arxiv link][project link][code|official][Tsinghua University + Shanghai Qi Zhi Institute + Shanghai AI Lab + Fudan University + Spirit AI; Yang Gao]

• BitVLA(arxiv2025.06) BitVLA: 1-bit Vision-Language-Action Models for Robotics Manipulation [arxiv link][code|official][Chinese Academy of Sciences + University of Chinese Academy of Sciences]

• BridgeVLA(arxiv2025.06) BridgeVLA: Input-Output Alignment for Efficient 3D Manipulation Learning with Vision-Language Models [arxiv link][project link][code|official][CASIA + Bytedance Seed + UCAS + FiveAges + NJU; Tieniu Tan]

• SmolVLA(arxiv2025.06) SmolVLA: A Vision-Language-Action Model for Affordable and Efficient Robotics [arxiv link][huggingface link][code|official][Hugging Face + Sorbonne University]

• TGRPO(arxiv2025.06) TGRPO :Fine-tuning Vision-Language-Action Model via Trajectory-wise Group Relative Policy Optimization [arxiv link][code|official][Jilin University]

• RationalVLA(arxiv2025.06) RationalVLA: A Rational Vision-Language-Action Model with Dual System [arxiv link][project link][The Hong Kong University of Science and Technology (Guangzhou) + Shanghai Jiao Tong University]

• DexVLG(ICCV2025)(arxiv2025.07) DexVLG: Dexterous Vision-Language-Grasp Model at Scale [arxiv link][project link][code|official][BAAI + Galbot + THU + PKU + CASIA + SJTU + EIT; He Wang]

• AC-DiT(arxiv2025.07) AC-DiT: Adaptive Coordination Diffusion Transformer for Mobile Manipulation [arxiv link][project link][code|official][Peking University + Nanjing University (NJU) + The Chinese University of Hong Kong (CUHK) + Beijing Academy of Artificial Intelligence (BAAI)]

• villa-X(arxiv2025.07) villa-X: Enhancing Latent Action Modeling in Vision-Language-Action Models [arxiv link][project link][code|official][Microsoft Research + Tsinghua University + Wuhan University + Hong Kong University of Science and Technology + Nanjing University]

• ReconVLA(AAAI2026)(arxiv2025.08) ReconVLA: Reconstructive Vision-Language-Action Model as Effective Robot Perceiver [arxiv link][project link][code|official][The Hong Kong University of Science and Technology (Guangzhou) + Westlake University + Zhejiang University + Monash University]

• MemoryVLA(arxiv2025.08) MemoryVLA: Perceptual-Cognitive Memory in Vision-Language-Action Models for Robotic Manipulation [arxiv link][project link][Tsinghua University + Dexmal + MEGVII Technology + Tianjin University + Harbin Institute of Technology + StepFun; Gao Huang]

• TRANSIC(CoRL2024)(arxiv2024.05) TRANSIC: Sim-to-Real Policy Transfer by Learning from Online Correction [openreview link][arxiv link][project link][code|official][Stanford University; Jiajun Wu + Li Fei-Fei]

• Manipulate-Anything(CoRL2024)(arxiv2024.06) Manipulate-Anything: Automating Real-World Robots using Vision-Language Models [openreview link][arxiv link][project link][code|official][University of Washington + NVIDIA + Allen Institute for Artifical Intelligence + Universidad Católica San Pablo; Dieter Fox][It has an Error Recovery module]

• SpatialBot(ICRA2025)(arxiv2024.06) SpatialBot: Precise Spatial Understanding with Vision Language Models [arxiv link][SpatialBench link][weixin blog][code|official][SJTU + Stanford + BAAI + PKU + Oxford + SEU]

• 👍RACER(ICRA2025)(arxiv2024.09) RACER: Rich Language-Guided Failure Recovery Policies for Imitation Learning [arxiv link][project link][code|official][University of Michigan]

• AHA(ICLR2025)(arxiv2024.10) AHA: A Vision-Language-Model for Detecting and Reasoning over Failures in Robotic Manipulation [openreview link][arxiv link][project link][code|official][NVIDIA + University of Washington + Universidad Católica San Pablo + MIT + Nanyang Technological University + Allen Institute for Artificial Intelligence; Dieter Fox]

• Spatially-Visual-Perception(arxiv2024.11) Spatially Visual Perception for End-to-End Robotic Learning [arxiv link][ZhiCheng AI + Peking University + Harvard University + Zhejiang University]

• RoboSpatial(CVPR2025, Oral)(arxiv2024.11) RoboSpatial: Teaching Spatial Understanding to 2D and 3D Vision-Language Models for Robotics [paper link][arxiv link][The Ohio State University + NVIDIA]

• 👍Code-as-Monitor(CVPR2025)(arxiv2024.12) Code-as-Monitor: Constraint-aware Visual Programming for Reactive and Proactive Robotic Failure Detection [paper link][arxiv link][project link][Beihang University + Peking University + Beijing Academy of Artificial Intelligence + GalBot; He Wang]

• RoboMD(arxiv2024.12) From Mystery to Mastery: Failure Diagnosis for Improving Manipulation Policies [arxiv link][project link][code|official][Arizona State University + University of Washington + NVIDIA; Dieter Fox]

• 👍FOREWARN(RSS2025)(arxiv2025.02) From Foresight to Forethought: VLM-In-the-Loop Policy Steering via Latent Alignment [paper link][arxiv link][project link][Carnegie Mellon University + UC Berkeley]

• 👍FAIL-Detect(RSS2025)(arxiv2025.03) Can We Detect Failures Without Failure Data? Uncertainty-Aware Runtime Failure Detection for Imitation Learning Policies [paper link][arxiv link][project link][code|official][Toyota Research Institute (TRI) + Woven by Toyota (WbyT)]

• AmbResVLM(arxiv2025.04) Robotic Task Ambiguity Resolution via Natural Language Interaction [arxiv link][project link][code|official][University of Freiburg + Toyota Motor Europe]

• RoboFAC(arxiv2025.05) RoboFAC: A Comprehensive Framework for Robotic Failure Analysis and Correction [arxiv link][project link][code|official][Shanghai Jiao Tong University + Xiamen University + Harbin Institute of Technology, Shenzhen]

• UNISafe(arxiv2025.05) Uncertainty-aware Latent Safety Filters for Avoiding Out-of-Distribution Failures [arxiv link][project link][code|official][Carnegie Mellon University]

• SAFE(arxiv2025.06) SAFE: Multitask Failure Detection for Vision-Language-Action Models [arxiv link][project link][code|official][University of Toronto (UofT) + UofT Robotics Institute + Vector Institute + Toyota Research Institute (TRI)][It introduces the multitask failure detection problem for VLA models, and propose SAFE, a failure detector that can detect failures for unseen tasks zero-shot and achieve state-of-the-art performance.]

• RoboRefer(NeurIPS2025)(arxiv2025.06) RoboRefer: Towards Spatial Referring with Reasoning in Vision-Language Models for Robotics [arxiv link][project link][code|official][Beihang University + Peking University + Beijing Academy of Artificial Intelligence]

• Dyna-Nonprehensile(IJRR1999) Dynamic Nonprehensile Manipulation: Controllability, Planning, and Experiments [paper link]

• Progress-Nonprehensile(IJRR1999) Progress in Nonprehensile Manipulation [paper link]

• Pushing-Skill(IJRR2019) Pushing revisited: Differential flatness, trajectory planning, and stabilization [paper link]

• MPC-LMS(IJRR2020) Reactive planar non-prehensile manipulation with hybrid model predictive control [paper link][Model Predictive Controller with Learned Mode Scheduling (MPC-LMS)]

• Pregrasp-Manipulation(ICRA2020)(arxiv2020.02) Learning Pregrasp Manipulation of Objects from Ungraspable Poses [paper link][arxiv link][Tsinghua University + University of Edinburg]

• Bimanual-Stir-fry(RAL2022)(arxiv2022.05) Robot Cooking with Stir-fry: Bimanual Non-prehensile Manipulation of Semi-fluid Objects [paper link][arxiv link][The Chinese University of Hong Kong + Wuhan University + Idiap Research Institute]

• 👍Occluded-Grasping(CoRL2022 Oral)(arxiv2022.11) Learning to Grasp the Ungraspable with Emergent Extrinsic Dexterity [openreview link][paper link][arxiv link][project link][code|official][Robotics Institute, Carnegie Mellon University]

• Pivoting(ICRA2023)(arxiv2023.05) Learning Generalizable Pivoting Skills [paper link][arxiv link][UC Berkeley + Mitsubishi Electric Research Laboratories (MERL) + Rutgers University]

• 👍HACMan(CoRL2023 Oral)(arxiv2023.05) HACMan: Learning Hybrid Actor-Critic Maps for 6D Non-Prehensile Manipulation [paper link][openreview link][arxiv link][project link][code|official][CMU + Meta]

• ED-PMP(ICRA2024)(arxiv2023.10) Learning Extrinsic Dexterity with Parameterized Manipulation Primitives [paper link][arxiv link][project link][Örebro University]

• MRLM(RAL2024)(arxiv2023.07) Multi-Stage Reinforcement Learning for Non-Prehensile Manipulation [paper link][arxiv link][project link][School of Control Science and Engineering, Shandong University]

• 👍CORN(ICLR2024)(arxiv2024.03) CORN: Contact-based Object Representation for Nonprehensile Manipulation of General Unseen Objects [openreview link][arxiv link][project link][code|official][Korea Advanced Institute of Science and Technology (KAIST) + Kim Jaechul Graduate school of AI]

• ExtrinsicManipulation(IROS2024)(arxiv2024.04) One-Shot Transfer of Long-Horizon Extrinsic Manipulation Through Contact Retargeting [paper link][arxiv link][project link][code|official][Stanford University + NVIDIA]

• Tactile-Non-Prehensile(RSS2024)(arxiv2024.05) Tactile-Driven Non-Prehensile Object Manipulation via Extrinsic Contact Mode Control [paper link][arxiv link][project link][University of Michigan]

• HACMan++(RSS2024)(arxiv2024.07) HACMan++: Spatially-Grounded Motion Primitives for Manipulation [paper link][arxiv link][project link][code|official][Carnegie Mellon University + Meta AI]

• Bimanual-Nonprehensile-Mani(ICRA2025)(arxiv2024.09) In the Wild Ungraspable Object Picking with Bimanual Nonprehensile Manipulation [paper link][arxiv link][Stanford University + Toyota Research Institute (TRI)]

• Nonprehensile-Rearrangement(TRO2025)(arxiv2024.10) Object-Centric Kinodynamic Planning for Nonprehensile Robot Rearrangement Manipulation [paper link][arxiv link][Rice University + The AI Institute, Cambridge]

• HyDo(RAL2025)(arxiv2024.11) Enhancing Exploration with Diffusion Policies in Hybrid Off-Policy RL: Application to Non-Prehensile Manipulation [paper link][arxiv link][project link][Bosch Center for Artificial Intelligence(BCAI) + Karlsruhe Institute of Technology][Hybrid Diffusion Policy algorithm (HyDo)]

• LearnVEC(CoRL2024)(arxiv2024.12) Learning Visuotactile Estimation and Control for Non-prehensile Manipulation under Occlusions [openreview link][arxiv link][video link][The University of Edinburgh + The Alan Turing Institute]

• PBPF(TRO2025) Tracking and Control of Multiple Objects During Nonprehensile Manipulation in Clutter [paper link][pdf link][code|official][University of Leeds + American University of Beirut - Mediterraneo]

• Quasi-Static-Pushing(TRO2025) Quasi-Static Modeling and Controlling for Planar Pushing of Deformable Objects [paper link][SJTU; Hesheng Wang]

• Robust-Pushing(IJRR2025) Robust pushing: Exploiting quasi-static belief dynamics and contact-informed optimization [paper link][EPFL + University of Oxford]

• 👍COMBO-Grasp(arxiv2025.02) COMBO-Grasp: Learning Constraint-Based Manipulation for Bimanual Occluded Grasping [arxiv link][project link][Applied AI Lab, University of Oxford]

• 👍HAMNET(RSS2025)(arxiv2025.02) Hierarchical and Modular Network on Non-prehensile Manipulation in General Environments [paper link][arxiv link][project link][KAIST]

• Skill-RRT(arxiv2025.02) SPIN: distilling Skill-RRT for long-horizon prehensile and non-prehensile manipulation [arxiv link][project link][Korea Advanced Institute of Science and Technology (KAIST)]

• Mobile-Pushing(ICRA2025)(arxiv2025.02) Dynamic object goal pushing with mobile manipulators through model-free constrained reinforcement learning [arxiv link][HHCM lab, IIT, Genoa 16163, Italy + DIBRIS, University of Genoa, Genoa 16145, Italy + RSL, ETH Z ̈urich, Z ̈urich 8092, Switzerland + NVIDIA]

• DyWA(ICCV2025)(arxiv2025.03) DyWA: Dynamics-adaptive World Action Model for Generalizable Non-prehensile Manipulation [arxiv link][project link][code|official][Peking University + Galbot; He Wang]

• ExDex(arxiv2025.03) Dexterous Non-Prehensile Manipulation for Ungraspable Object via Extrinsic Dexterity [arxiv link][project link][PKU-PsiBot Joint Lab + Peking University]

• ProbabilisticPrehensilePushing(RAL2025)(arxiv2025.03) Pushing Everything Everywhere All At Once: Probabilistic Prehensile Pushing [paper link][arxiv link][project link][code|official][Division of Robotics, Perception and Learning (RPL), KTH]

• 👍PIN-WM(RSS2025)(arxiv2025.04) PIN-WM: Learning Physics-INformed World Models for Non-Prehensile Manipulation [paper link][arxiv link][project link][code|official][National University of Defense Technology + Wuhan University + Shenzhen University + Guangdong Laboratory of Artificial Intelligence and Digital Economy]

• ActivePusher(arxiv2025.06) ActivePusher: Active Learning and Planning with Residual Physics for Nonprehensile Manipulation [arxiv link][Worcester Polytechnic Institute]

• DexNoMa(year2025.06) DexNoMa: Learning Geometry-Aware Nonprehensile Dexterous Manipulation [openreview link][project link][University of Southern California]

• 👍ParticleFormer(arxiv2025.06) ParticleFormer: A 3D Point Cloud World Model for Multi-Object, Multi-Material Robotic Manipulation [arxiv link][project link][Stanford University + RAI Institute][ParticleFormer captures fine-grained multi-object interactions between rigid, deformable, and flexible materials, trained directly from real-world robot perception data without an elaborate scene reconstruction.][This work proposes to use World Model to address tasks Box Pushing and Rope Sweeping.]

• CVAE+RL(arxiv2025.07) Leveraging Extrinsic Dexterity for Occluded Grasping on Grasp Constraining Walls [arxiv link][University of California, Berkeley]

• CITO-RL(arxiv2025.08) Learning Pivoting Manipulation with Force and Vision Feedback Using Optimization-based Demonstrations [arxiv link][video link][Mitsubishi Electric Research Laboratories + Mitsubishi Electric]

• GD2P(arxiv2025.09) Learning Geometry-Aware Nonprehensile Pushing and Pulling with Dexterous Hands [arxiv link][project link][University of Southern California]

• Where2Act(ICCV2021)(arxiv2021.01) Where2Act: From Pixels to Actions for Articulated 3D Objects [paper link][arxiv link][project link][code|official][Stanford University + Facebook AI Research][It is based on the simulation SAPIEN]

• 👍DextAIRity(RSS2022, Best Systems Paper Award Finalist)(arxiv2022.03) DextAIRity: Deformable Manipulation Can be a Breeze [paper link][arxiv link][project link][code|official][Columbia University + Toyota Research Institute; Shuran Song]

• BimanualAssitDressing(TRO2024)(arxiv2023.01) Do You Need a Hand? a Bimanual Robotic Dressing Assistance Scheme [paper link][arxiv link][project link][University of York + Honda Research Institute + Cognitive Robotics, 3mE, TU Delft, the Netherlands]

• LLM-AOM(ICRA2024)(arxiv2023.11) Kinematic-aware Prompting for Generalizable Articulated Object Manipulation with LLMs [arxiv link][project link][code|official][Renmin University of China + Shanghai Artificial Intelligence Laboratory + Northwestern Polytechnical University][the Demonstration Collection scripts on Isaac gym]

• RPMArt(IROS2024)(arxiv2024.03) RPMArt: Towards Robust Perception and Manipulation for Articulated Objects [arxiv link][project link][code|official][Shanghai Jiao Tong University + Stanford University + Hefei University of Technology; Cewu Lu]

• BimanualTwist(CoRL2024)(arxiv2024.03) Twisting Lids Off with Two Hands [openreview link][paper link][arxiv link][project link][code|official][UC Berkeley; Pieter Abbeel][It used two multi-fingered robot hands]

• DigitalTwinArt(CVPR2024)(arxiv2024.04) Neural Implicit Representation for Building Digital Twins of Unknown Articulated Objects [paper link][arxiv link][project link][code|official][NVIDIA + Stanford University]

• 👍A3VLM(CoRL2024)(arxiv2024.06) A3VLM: Actionable Articulation-Aware Vision Language Model [openreview link][arxiv link][code|official][SJTU + Shanghai AI Lab + Rutgers University + Yuandao AI + PKU + CUHK MMLab]

• RSRD(CoRL2024, Oral)(arxiv2024.09) Robot See Robot Do: Imitating Articulated Object Manipulation with Monocular 4D Reconstruction [openreview link][arxiv link][project link][code|official][UC Berkeley]

• DexSim2Real2(arxiv2024.09) DexSim2Real2: Building Explicit World Model for Precise Articulated Object Dexterous Manipulation [arxiv link][project link][Tsinghua University + JD Explore Academy]

• AxisEst(arxiv2024.09) Articulated Object Manipulation using Online Axis Estimation with SAM2-Based Tracking [arxiv link][project link][code|official][University of Hong Kong + Shenzhen University + Shanghai Jiaotong University + Southern University of Science and Technology][Articulated Object Manipulation]

• UniAff(arxiv2024.09) UniAff: A Unified Representation of Affordances for Tool Usage and Articulation with Vision-Language Models [arxiv link][project link][SJTU + HKUST + NUS + Rutgers University; Cewu Lu]

• BimArt(CVPR2025)(arxiv2024.12) BimArt: A Unified Approach for the Synthesis of 3D Bimanual Interaction with Articulated Objects [paper link][arxiv link][project link][MPII + Google]

• RoboHanger(arxiv2024.12) RoboHanger: Learning Generalizable Robotic Hanger Insertion for Diverse Garments [arxiv link][project link][Peking University + GALBOT + Beijing Academy of Artificial Intelligence+ He Wang]

• GAPS(ICASSP2025)(arxiv2024.12) Generalizable Articulated Object Perception with Superpoints [paper link][arxiv link][Shanghai Jiao Tong University + National University of Singapore + University of Science and Technology of China + Hefei University of Technology + National Institute of Technology; Cewu Lu]

• AdaManip(ICLR2025)(arxiv2025.02) AdaManip: Adaptive Articulated Object Manipulation Environments and Policy Learning [openreview link][arxiv link][project link][code|official][Peking University + Beijing University of Posts and Telecommunications; Hao Dong]

• Watch-Less-Feel-More(ICRA2025)(arxiv2025.02) Watch Less, Feel More: Sim-to-Real RL for Generalizable Articulated Object Manipulation via Motion Adaptation and Impedance Control [arxiv link][project link][Peking University + Galbot; He Wang]

• UniClothDiff(arxiv2025.03) Diffusion Dynamics Models with Generative State Estimation for Cloth Manipulation [arxiv link][University of California San Diego + Hillbot; Hao Su]

• CoDA(arxiv2025.05) CoDA: Coordinated Diffusion Noise Optimization for Whole-Body Manipulation of Articulated Objects [arxiv link][project link][code|official][The University of Hong Kong + Zhejiang University]

• Infinigen-Sim(arxiv2025.05) Infinigen-Sim: Procedural Generation of Articulated Simulation Assets [arxiv link][project link][code|official][Princeton University + McGill University + Cisco; Jia Deng][Infinigen-Sim is a toolkit for procedurally generating high-quality articulated assets for robotics simulation.]

• DexGarmentLab(arxiv2025.05) DexGarmentLab: Dexterous Garment Manipulation Environment with Generalizable Policy [arxiv link][project link][code|official][Peking University + University of California, Berkeley; Pieter Abbeel + Hao Dong]

• ArtVIP(arxiv2025.06) ArtVIP: Articulated Digital Assets of Visual Realism, Modular Interaction, and Physical Fidelity for Robot Learning [arxiv link][project link][code|official][Beijing Innovation Center of Humanoid Robotics + Beijing Institute of Architectural Design]

• PhysRig(ICCV2025)(arxiv2025.06) PhysRig: Differentiable Physics-Based Skinning and Rigging Framework for Realistic Articulated Object Modeling [paper link][arxiv link][project link][code|official][University of Illinois Urbana-Champaign + Stability AI]

• DreamArt(arxiv2025.07) DreamArt: Generating Interactable Articulated Objects from a Single Image [arxiv link][project link][Peking University + Tsinghua University + BIGAI; Siyuan Huang][DreamArt is capable of synthesizing articulated objects from a single image]

• Handover-Sim2Sim(CVPR2023, highlight)(arxiv2023.03) Learning Human-to-Robot Handovers from Point Clouds [paper link][arxiv link][project link][code|official][ETH Zurich + NVIDIA + University of Washington; Dieter Fox]

• SynH2R(ICRA2024)(arxiv2023.11) SynH2R: Synthesizing Hand-Object Motions for Learning Human-to-Robot Handovers [paper link][arxiv link][project link][Department of Computer Science, ETH Zurich + NVIDIA]

• 👍GenH2R(CVPR2024)(arxiv2024.01) GenH2R: Learning Generalizable Human-to-Robot Handover via Scalable Simulation Demonstration and Imitation [paper link][arxiv link][project link][code|official][Tsinghua University + Shanghai Artificial Intelligence Laboratory + Shanghai Qi Zhi Institute]

• CyberDemo(CVPR2024)(arxiv2024.02) CyberDemo: Augmenting Simulated Human Demonstration for Real-World Dexterous Manipulation [arxiv link][project link][UC San Diego + USC; related to Xiaolong Wang group; using the Allegro Hand to conduct their real robot experiments.]

• CORE4D(CVPR2025)(arxiv2024.06) CORE4D: A 4D Human-Object-Human Interaction Dataset for Collaborative Object REarrangement [paper link][arxiv link][project link][code|official][1Tsinghua University + Shanghai Qi Zhi Institute + Shanghai Artificial Intelligence Laboratory + Beijing University of Posts and Telecommunications]

• HumanVLA(NIPS2024)(arxiv2024.06) HumanVLA: Towards Vision-Language Directed Object Rearrangement by Physical Humanoid [openreview link][arxiv link][code|official][Shanghai Jiao Tong University + Tencent Robotics X; Yong-Lu Li + Cewu Lu]

• GAAF-DEX(TNNLS2025)(arxiv2024.07) Learning Granularity-Aware Affordances from Human-Object Interaction for Tool-Based Functional Dexterous Grasping [paper link][arxiv link][code|official][Hunan University; Yaonan Wang]

• MobileH2R(CVPR2025)(arxiv2025.01) MobileH2R: Learning Generalizable Human to Mobile Robot Handover Exclusively from Scalable and Diverse Synthetic Data [paper link][arxiv link][project link][Tsinghua University + Galbot + Peking University + Shanghai Artificial Intelligence Laboratory + Shanghai Qi Zhi Institute; He Wang]

• InterMimic(CVPR2025 highlight)(arxiv2025.02) InterMimic: Towards Universal Whole-Body Control for Physics-Based Human-Object Interactions [arxiv link][project link][code|official][University of Illinois Urbana Champaign + Electronic Arts]

• 👍RHINO(arxiv2025.02) RHINO: Learning Real-Time Humanoid-Human-Object Interaction from Human Demonstrations [arxiv link][project link][code|official][Shanghai Jiao Tong University; Weinan Zhang]

• Dyn-R2H-Handover(ICRA2025)(arxiv2025.02) Learning-based Dynamic Robot-to-Human Handover [arxiv link][project link][Korea University + Queen's University]

• Sketch-to-Skill(arxiv2025.03) Sketch-to-Skill: Bootstrapping Robot Learning with Human Drawn Trajectory Sketches [arxiv link][openreview link (ICLR2025, rejected)][University of Maryland, College Park, Maryland]

• ADC-Robot(arxiv2025.03) Adversarial Data Collection: Human-Collaborative Perturbations for Efficient and Robust Robotic Imitation Learning [arxiv link][project link][code|official][Shanghai Jiao Tong University + CUHK + Agibot + Beihang University]

• 👍RoboCopilot(arxiv2025.03) RoboCopilot: Human-in-the-loop Interactive Imitation Learning for Robot Manipulation [arxiv link][University of California, Berkeley; Pieter Abbeel]

• HAPO(NIPS2025)(arxiv2025.06) Robotic Policy Learning via Human-assisted Action Preference Optimization [arxiv link][project link][Renmin University of China + ByteDance Seed]

• DexH2R(ICCV2025)(arxiv2025.06) DexH2R: A Benchmark for Dynamic Dexterous Grasping in Human-to-Robot Handover [arxiv link][project link][code|official][ShanghaiTech University + The University of Hong Kong]

• H2RH-SGS(arxiv2025.07) Learning human-to-robot handovers through 3D scene reconstruction [arxiv link][project link][code|official][Queen Mary University of London + Idiap Research Institute + École Polytechnique Fédérale de Lausanne]

• HERMES(arxiv2025.08) HERMES: Human-to-Robot Embodied Learning from Multi-Source Motion Data for Mobile Dexterous Manipulation [arxiv link][project link][Tsinghua University + Shanghai Qi Zhi Institute + Peking University; Huazhe Xu]

• 👍AgentWorld(CoRL2025)(arxiv2025.08) AgentWorld: An Interactive Simulation Platform for Scene Construction and Mobile Robotic Manipulation [arxiv link][project link][code|official][Tencent Robotics X + Shanghai Jiao Tong University; Cewu Lu]

• MICoBot(arxiv2025.08) Mixed-Initiative Dialog for Human-Robot Collaborative Manipulation [arxiv link][project link][UT Austin + Stanford]

• MimicPlay(CoRL2023 Oral)(arxiv2023.02) MimicPlay: Long-Horizon Imitation Learning by Watching Human Play [openreview link][paper link][arxiv link][project link][code|official][Stanford + NVIDIA + Georgia Tech + UT Austin + Caltech, by Stanford Fei-Fei Li]

• VRB(CVPR2023)(arxiv2023.04) Affordances from Human Videos as a Versatile Representation for Robotics [paper link][arxiv link][project link][code|official][Carnegie Mellon University + Meta AI]

• Giving-Robots-a-Hand(arxiv2023.07) Giving Robots a Hand: Learning Generalizable Manipulation with Eye-in-Hand Human Video Demonstrations [arxiv link][project link][Stanford University]

• HOPMan(ICRA2024, Best Paper in Robot Manipulation Finalist)(arxiv2023.12) Towards Generalizable Zero-Shot Manipulation via Translating Human Interaction Plans [paper link][arxiv link][project link][Carnegie Mellon University + Meta AI]

• 👍GeneralFlow/GFlow(CoRL2024)(arxiv2024.01) General Flow as Foundation Affordance for Scalable Robot Learning [openreview link][paper link][arxiv link][project link][code|official][Tsinghua University + Shanghai Artificial Intelligence Laboratory + Shanghai Qi Zhi Institute; Yang Gao][This method can learn from few-shot human video demonstrations using the General Flow as Affordance representation.]

• 👍UMI(RSS2024, Best Systems Paper Award Finalist)(arxiv2024.02) Universal Manipulation Interface: In-The-Wild Robot Teaching Without In-The-Wild Robots [paper link][arxiv link][project link][code|official][Stanford University + Columbia University + Toyota Research Insititute; Shuran Song]

• DITTO(IROS2024)(arxiv2024.03) DITTO: Demonstration Imitation by Trajectory Transformation [paper link][arxiv link][project link][University of Freiburg, Germany][Learning from action labels free human videos]

• ViViDex(ICRA2025)(arxiv2024.04) ViViDex: Learning Vision-based Dexterous Manipulation from Human Videos [arxiv link][project link][Inria + Mohamed bin Zayed University of Artificial Intelligence]

• ORION(CoRL2024 Workshop)(arxiv2024.05) Vision-based Manipulation from Single Human Video with Open-World Object Graphs [openreview link][arxiv link][project link][The University of Texas at Austin + Sony AI; Yuke Zhu][Learning from action labels free human videos][We investigate the problem of imitating robot manipulation from a single human video in the open-world setting, where a robot must learn to manipulate novel objects from one video demonstration.]

• 👍ScrewMimic(RSS2024, Outstanding Student Paper Award Finalist)(arxiv2024.05) ScrewMimic: Bimanual Imitation from Human Videos with Screw Space Projection [arxiv link][project link][code|official][The University of Texas at Austin]

• HumanRobotAlign(CVPR2025)(arxiv2024.06) Mitigating the Human-Robot Domain Discrepancy in Visual Pre-training for Robotic Manipulation [paper link][arxiv link][project link][HKUST-GZ]

• HRP(RSS2024)(arxiv2024.07) HRP: Human Affordances for Robotic Pre-Training [paper link][arxiv link][project link][code|official][Carnegie Mellon University]

• 👍R+X(ICRA2025)(arxiv2024.07) R+X: Retrieval and Execution from Everyday Human Videos [arxiv link][project link][The Robot Learning Lab at Imperial College London; Edward Johns][Learning from action labels free human videos]

• TieBot(CoRL2024, Oral)(arxiv2024.07) TieBot: Learning to Knot a Tie from Visual Demonstration through a Real-to-Sim-to-Real Approach [openreview link][arxiv link][project link][National University of Singapore + Shanghai Jiao Tong University + Nanjing University; Cewu Lu][Learning from action labels free human videos]

• Gen2Act(CoRL2025)(arxiv2024.09) Gen2Act: Human Video Generation in Novel Scenarios enables Generalizable Robot Manipulation [paper link][arxiv link][project link][Google DeepMind + Carnegie Mellon University + Stanford University]

• ObjDex(CoRL2024) Object-Centric Dexterous Manipulation from Human Motion Data [openreview link][project link][Stanford University + Peking University]

• BiDexHD(arxiv2024.10) Learning Diverse Bimanual Dexterous Manipulation Skills from Human Demonstrations [arxiv link][project link][ICLR2025 rejected][Peking University][It is based on the TACO Dataset and Isaac Gym]

• SeeDo(IROS2025)(arxiv2024.10) VLM See, Robot Do: Human Demo Video to Robot Action Plan via Vision Language Model [arxiv link][project link][code|official][New York University]

• SBAMs(Humanoids 2024)(arxiv2024.10) Learning Spatial Bimanual Action Models Based on Affordance Regions and Human Demonstrations [arxiv link][Karlsruhe Institute of Technology][Spatial Bimanual Action Models]

• HuDOR(RSS2025 Workshop)(arxiv2024.10) HuDOR: Bridging the Human to Robot Dexterity Gap through Object-Oriented Rewards [openreview link][arxiv link][project link][New York University]

• 👍EgoMimic(CoRL2024 Workshop)(arxiv2024.10) EgoMimic: Scaling Imitation Learning via Egocentric Video [openreview link][arxiv link][project link][code|official][Georgia Tech + Stanford University][bimanual manipulation, learning from human videos]

• OKAMI(CoRL2024, Oral)(arxiv2024.10) OKAMI: Teaching Humanoid Robots Manipulation Skills through Single Video Imitation [openreview link][arxiv link][project link][UT Austin + NVIDIA Research; Yuke Zhu][It enables a humanoid robot to imitate manipulation skills from a single human video demonstration.]

• 👍LAPA(ICLR2025)(arxiv2024.10) Latent Action Pretraining from Videos [openreview link][arxiv link][project link][code|official][KAIST + University of Washington + Microsoft Research + NVIDIA + Allen Institute for AI; Dieter Fox]

• DexH2R(arxiv2024.11) DexH2R: Task-oriented Dexterous Manipulation from Human to Robots [arxiv link][University of California, Berkeley]

• 👍MT-Policy(ICRA2025)(arxiv2025.01) Motion Tracks: A Unified Representation for Human-Robot Transfer in Few-Shot Imitation Learning [arxiv link][project link][Cornell University + Stanford University]

• MocapRobot(arxiv2025.01) Learning to Transfer Human Hand Skills for Robot Manipulations [arxiv link][project link][Seoul National University + Carnegie Mellon University]

• 👍👍FUNCTO(arxiv2025.02) FUNCTO: Function-Centric One-Shot Imitation Learning for Tool Manipulation [arxiv link][project link][Southern University of Science and Technology + National University of Singapore][A key challenge lies in establishing functional correspondences between demonstration and test tools]

• Human2Robot(arxiv2025.02) Human2Robot: Learning Robot Actions from Paired Human-Robot Videos [arxiv link][Fudan University]

• 2HandedAfforder(ICCV2025)(RSS2025 Workshop)(arxiv2025.03) 2HandedAfforder: Learning Precise Actionable Bimanual Affordances from Human Videos [paper link][arxiv link][project link][code|official][PEARL Lab, TU Darmstadt, Germany]

• VidBot(CVPR2025)(arxiv2025.03) VidBot: Learning Generalizable 3D Actions from In-the-Wild 2D Human Videos for Zero-Shot Robotic Manipulation [paper link][arxiv link][project link][code|official][Technical University of Munich + ETH Zürich + Microsoft]

• FreePose(ICLR2025)(arxiv2025.03) 6D Object Pose Tracking in Internet Videos for Robotic Manipulation [arxiv link][project link][code|official][Czech Technical University in Prague + H Company]

• 👍HumanoidPolicy(arxiv2025.03) Humanoid Policy ~ Human Policy [arxiv link][project link][code|official][UC San Diego + CMU + University of Washington + MIT + Apple; Xiaolong Wang]

• Human2Sim2Robot(CoRL2025)(arxiv2025.04) Crossing the Human-Robot Embodiment Gap with Sim-to-Real RL using One Human Demonstration [arxiv link][project link][code|official][Stanford University]

• 👍Tool-as-Interface(CoRL2025)(arxiv2025.04) Tool-as-Interface: Learning Robot Policies from Observing Human Tool Use [arxiv link][project link][code|official][University of Illinois Urbana-Champaign + UT Austin + Columbia University]

• H2R(arxiv2025.05) H2R: A Human-to-Robot Data Augmentation for Robot Pre-training from Videos [arxiv link][Peking University + University of Washington]

• 👍👍DexUMI(CoRL2025, Best Paper Final List)(arxiv2025.05) DexUMI: Using Human Hand as the Universal Manipulation Interface for Dexterous Manipulation [arxiv link][project link][code|official][Stanford University + Columbia University + J.P. Morgan AI Research + Carnegie Mellon University + NVIDIA]

• 👍HANDRetrieval(arxiv2025.05) HAND Me the Data: Fast Robot Adaptation via Hand Path Retrieval [arxiv link][project link][code|official][University of Southern California][We introduce HAND, a simple and time-efficient method for teaching robots manipulation tasks through human hand demonstrations.]

• 👍👍BLADE(CoRL2024)(arxiv2025.05) Learning Compositional Behaviors from Demonstration and Language [openreview link][arxiv link][project link][Stanford University + MIT; Jiajun Wu]

• GLOVER++(arxiv2025.05) GLOVER++: Unleashing the Potential of Affordance Learning from Human Behaviors for Robotic Manipulation [arxiv link][project link][code|official][HKUST (GZ) + HKUST]

• PDCP(arxiv2025.05) Learning Generalizable Robot Policy with Human Demonstration Video as a Prompt [arxiv link][Tsinghua University + Shanghai Qi Zhi Institute + RobotEra]

• 3DMF(arxiv2025.06) Object-centric 3D Motion Field for Robot Learning from Human Videos [arxiv link][project link][UC Berkeley EECS + Google DeepMind; Pieter Abbeel]

• DemoDiffusion(arxiv2025.06) DemoDiffusion: One-Shot Human Imitation using pre-trained Diffusion Policy [arxiv link][project link][code|official][Carnegie Mellon University]

• RwoR(IROS2025)(arxiv2025.07) RwoR: Generating Robot Demonstrations from Human Hand Collection for Policy Learning without Robot [arxiv link][project link][Peking University + Tencent Robotics X Laboratory; Hao Dong]

• 👍EgoVLA(arxiv2025.07) EgoVLA: Learning Vision-Language-Action Models from Egocentric Human Videos [arxiv link][project link][code|official][UC San Diego + UIUC + MIT + NVIDIA; Xiaolong Wang]

• 👍Being-H0(arxiv2025.07) Being-H0: Vision-Language-Action Pretraining from Large-Scale Human Videos [arxiv link][project link][code|official][Peking University + Renmin University of China + BeingBeyond]

• 👍FMimic(IJRR2025)(arxiv2025.07) FMimic: Foundation Models are Fine-grained Action Learners from Human Videos [arxiv link][project link][Beijing Institute of Technology + The University of Hong Kong]

• MimicFunc(CoRL2025)(arxiv2025.08) MimicFunc: Imitating Tool Manipulation from a Single Human Video via Functional Correspondence [arxiv link][project link][code|official][Southern University of Science and Technology + National University of Singapore][Tool Manipulation, Imitation from Human Video]

• ImMimic(CoRL2025)(2025.09) ImMimic: Cross-Domain Imitation from Human Videos via Mapping and Interpolation [openreview link][arxiv link][project link][Georgia Institute of Technology]

• MimicDroid(arxiv2025.09) MimicDroid: In-Context Learning for Humanoid Robot Manipulation from Human Play Videos [arxiv link][project link][code|official][The University of Texas at Austin + Amazon Consumer Robotics + NVIDIA; Yuke Zhu]

• HDMI(arxiv2025.09) HDMI: Learning Interactive Humanoid Whole-Body Control from Human Videos [arxiv link][project link][code|official][Carnegie Mellon University]

• 👍MobileALOHA(CoRL2024)(arxiv2024.01) Mobile ALOHA: Learning Bimanual Mobile Manipulation with Low-Cost Whole-Body Teleoperation [openreview link][paper link][arxiv link][project link][code|official][Stanford University; Chelsea Finn]

• BRMData(arxiv2024.05) Empowering Embodied Manipulation: A Bimanual-Mobile Robot Manipulation Dataset for Household Tasks [arxiv link][project link][dataset link][JD Explore Academy][It proposed BRMData, a Bimanual-mobile Robot Manipulation Dataset specifically designed for household applications.]

• BiGym(CoRL2024)(arxiv2024.07) BiGym: A Demo-Driven Mobile Bi-Manual Manipulation Benchmark [openreview link][arxiv link][project link][code|official][Dyson Robot Learning Lab + Stephen James][Mujoco + UniTree H1][BiGym is a new benchmark and learning environment for mobile bi-manual demo-driven robotic manipulation.]

• 👍FlyingHand(RSS2025)(arxiv2024.07) Flying Hand: End-Effector-Centric Framework for Versatile Aerial Manipulation Teleoperation and Policy Learning [arxiv link][project link][Carnegie Mellon University + Pennsylvania State University]

• DIAL-MPC(ICRA2025, Best Paper Finalist)(arxiv2024.09) Full-Order Sampling-Based MPC for Torque-Level Locomotion Control via Diffusion-Style Annealing [arxiv link][project link][code|official][Carnegie Mellon University][DIAL-MPC: Diffusion-Inspired Annealing For Legged MPC]

• Catch_It(ICRA2025)(arxiv2024.09) Catch It! Learning to Catch in Flight with Mobile Dexterous Hands [arxiv link][project link][code|official][Shanghai Qi Zhi Institute + Tsinghua University + Shanghai AI Lab + Georgia Institute of Technology + Stanford University]

• AV-ALOHA(arxiv2024.09) Active Vision Might Be All You Need: Exploring Active Vision in Bimanual Robotic Manipulation [arxiv link][project link][code|official][University of California, Berkeley + University of California, Davis]

• BUMBLE(ICRA2025)(arxiv2024.10) BUMBLE: Unifying Reasoning and Acting with Vision-Language Models for Building-wide Mobile Manipulation [arxiv link][project link][code|official][The University of Texas at Austin; Yuke Zhu]

• 👍iDP3(arxiv2024.10) Generalizable Humanoid Manipulation with Improved 3D Diffusion Policies [arxiv link][project link][code|official][Stanford University + Simon Fraser University + UPenn + UIUC + CMU; Jiajun Wu]

• PIM(ICRA2025)(arxiv2024.11) Learning Humanoid Locomotion with Perceptive Internal Model [arxiv link][project link][code|official][Shanghai AI Laboratory + The University of Hong Kong + Zhejiang University + Shanghai Jiao Tong University; Ping Luo + Jiangmiao Pang]

• TidyBot++(CoRL2024)(arxiv2024.12) TidyBot++: An Open-Source Holonomic Mobile Manipulator for Robot Learning [openreview link][arxiv link][project link][code|official][Princeton University + Stanford University + Dexterity; Shuran Song]

• RoboMatrix(arxiv2024.12) RoboMatrix: A Skill-centric Hierarchical Framework for Scalable Robot Task Planning and Execution in Open-World [arxiv link][project link][code|official][weixin blog][Waseda University + Beijing Institute of Technology + Megvii Technology + The Chinese University of Hong Kong][Mobile Manipulation]

• HOMIE(arxiv2025.02) HOMIE: Humanoid Loco-Manipulation with Isomorphic Exoskeleton Cockpit [arxiv link][project link][code|official][Shanghai Artificial Intelligence Laboratory + The Chinese University of Hong Kong; Dahua Lin + Jiangmiao Pang]

• 👍Being-0(arxiv2025.03) Being-0: A Humanoid Robotic Agent with Vision-Language Models and Modular Skills [arxiv link][project link][code|official][PKU + BAAI + Being]

• 👍AhaRobot(arxiv2025.03) AhaRobot: A Low-Cost Open-Source Bimanual Mobile Manipulator for Embodied AI [arxiv link][project link][code|official][Tianjin University][A new Robot Platform named AhaRobot]

• MoMa-Kitchen(arxiv2025.03) MoMa-Kitchen: A 100K+ Benchmark for Affordance-Grounded Last-Mile Navigation in Mobile Manipulation [arxiv link][project link][code|official][Fudan University + Shanghai AI Laboratory + University of Science and Technology of China + Northwestern Polytechnical University + TeleAI, China Telecom Corp Ltd; Xuelong Li]

• MoManipVLA(CVPR2025)(arxiv2025.03) MoManipVLA: Transferring Vision-language-action Models for General Mobile Manipulation [arxiv link][project link][Beijing University of Posts and Telecommunications + Nanyang Technological University + Tsinghua University]

• 👍BRS(arxiv2025.03) BEHAVIOR Robot Suite: Streamlining Real-World Whole-Body Manipulation for Everyday Household Activities [arxiv link][project link][code|official][Stanford University; Shuran Song + Jiajun Wu + Li Fei-Fei][The hardware is named JoyLo: Joy-Con on Low-Cost Kinematic-Twin Arms; The model is named WB-VIMA: Whole-Body VisuoMotor Attention Policy]

• 👍SketchInterfacePoC(HRI2025)(arxiv2025.05) Sketch Interface for Teleoperation of Mobile Manipulator to Enable Intuitive and Intended Operation: A Proof of Concept [arxiv link][project link][Frontier Research Center, Toyota Motor Corporation + Aichi Institute of Technology]

• 👍Mobi-pi(CoRL2025)(arxiv2025.05) Mobi-pi: Mobilizing Your Robot Learning Policy [arxiv link][project link][code|official][Stanford University + Toyota Research Institute + University of Cambridge]

• UniFP(arxiv2025.05) Learning Unified Force and Position Control for Legged Loco-Manipulation [arxiv link][project link][BIGAI + UniTree Robotics + Beijing University of Posts and Telecommunications]

• DribbleMaster(arxiv2025.05) Dribble Master: Learning Agile Humanoid Dribbling Through Legged Locomotion [arxiv link][project link][Tsinghua University + Stanford University]

• AMO(RSS2025)(arxiv2025.05) AMO: Adaptive Motion Optimization for Hyper-Dexterous Humanoid Whole-Body Control [paper link][arxiv link][project link][code|official][UC San Diego; Xiaolong Wang]

• R2S2 / OpenWBT(arxiv2025.05) Unleashing Humanoid Reaching Potential via Real-world-Ready Skill Space [arxiv link][project link][code|official][Tsinghua University + Peking University + Galbot + Shanghai AI Laboratory + Shanghai Qi Zhi Institute + Nanjing University + Tongji University; He Wang]

• MAPPO(arxiv2025.05) Toward Real-World Cooperative and Competitive Soccer with Quadrupedal Robot Teams [arxiv link][UC Berkeley + Tsinghua University + Zhejiang University + Shanghai Qi Zhi Institute]

• EgoZero(arxiv2025.05) EgoZero: Robot Learning from Smart Glasses [arxiv link][project link][code|official][New York University + UC Berkeley; Pieter Abbeel]

• 👍ViA(CoRL2025)(arxiv2025.06) Vision in Action: Learning Active Perception from Human Demonstrations [arxiv link][project link][code|official][Stanford; Shuran Song]

• 👍❤MSLMaps(IJRR2025)(arxiv2025.06) Multimodal Spatial Language Maps for Robot Navigation and Manipulation [arxiv link][project link][code|official][University of Technology Nuremberg + UC Berkeley + Google Research]

• SLAC(arxiv2025.06) SLAC: Simulation-Pretrained Latent Action Space for Whole-Body Real-World RL [arxiv link][project link][The University of Texas at Austin + Sony AI + Amazon]

• ReLIC(arxiv2025.06) Versatile Loco-Manipulation through Flexible Interlimb Coordination [arxiv link][project link][RAI Institute + University of California, Berkeley + Cornell University]

• SkillBlender(arxiv2025.06) SkillBlender: Towards Versatile Humanoid Whole-Body Loco-Manipulation via Skill Blending [arxiv link][project link][University of Southern California + Stanford University + Peking University + University of California, Berkeley; Pieter Abbeel][SkillBlender performs versatile autonomous humanoid loco-manipulation tasks within different embodiments and environments, given only one or two intuitive reward terms.]

• 👍EyeRobot(arxiv2025.06) Eye, Robot: Learning to Look to Act with a BC-RL Perception-Action Loop [arxiv link][project link][code|official][UC Berkeley]

• 👍ActivePose(arxiv2025.09) ActivePose: Active 6D Object Pose Estimation and Tracking for Robotic Manipulation [arxiv link][Karlsruhe Institute of Technology, Germany + Shanghai Jiao Tong University, China + Istituto Italiano di Tecnologia, Italy + The Hong Kong Polytechnic University, Hong Kong + D-Robotics]

• GATO(TMLR2022)(arxiv2022.05) A Generalist Agent [openreview link]][arxiv link][offifial blog][code|not official][Deepmind]

• RoboCat(TMLR2023)(arxiv2023.06) RoboCat: A Self-Improving Generalist Agent for Robotic Manipulation [openreview link][arxiv link][Google DeepMind]

• LEO(ICML2024)(arxiv2023.11) An Embodied Generalist Agent in 3D World [arxiv link][project link][code|official][BIGAI + PKU + CMU + THU; on the simulator world]

• ReAd(arxiv2024.05) Towards Efficient LLM Grounding for Embodied Multi-Agent Collaboration [arxiv link][project link][THU + Shanghai AI Lab + Northwestern Polytechnical University + ZJU; Multi-Agent Collaboration][Reinforced Advantage feedback]

• SigmaAgent(arxiv2024.06) Contrastive Imitation Learning for Language-guided Multi-Task Robotic Manipulation [arxiv link][project link][HKUST-GZ][This work is partly based on the RVT; Sigma-Agent incorporates contrastive Imitation Learning (contrastive IL) modules to strengthen vision-language and current-future representations.]

• Make-An-Agent(NIPS2024)(arxiv2024.07) Make-An-Agent: A Generalizable Policy Network Generator with Behavior-Prompted Diffusion [arxiv link][project link][code|official][University of Maryland + College Park + Tsinghua University, IIIS + UC San Diego; Huazhe Xu]

• Magma(CVPR2025)(arxiv2025.02) Magma: A Foundation Model for Multimodal AI Agents [arxiv link][project link][code|official][Microsoft Research + University of Maryland + University of Wisconsin-Madison + KAIST + University of Washington]

• RoboBrain(CVPR2025)(arxiv2025.02) RoboBrain: A Unified Brain Model for Robotic Manipulation from Abstract to Concrete [arxiv link][project link][code|official][Peking University + Beijing Academy of Artificial Intelligence + Chinese Academy of Sciences + The University of Hong Kong + University of Chinese Academy of Sciences]

• AgenticRobot(arxiv2025.05) Agentic Robot: A Brain-Inspired Framework for Vision-Language-Action Models in Embodied Agents [arxiv link][project link][Jilin University + Harvard University + Massachusetts Institute of Technology + Huazhong University of Science and Technology + Southern University of Science and Technology + Lehigh University, Shanghai Jiao Tong University]

• OWMM-Agent(arxiv2025.06) OWMM-Agent: Open World Mobile Manipulation With Multi-modal Agentic Data Synthesis [arxiv link][code|official][Shanghai AI Laboratory + School of Computing, National University of Singapore + The Univeristy of Hongkong + Shanghai Jiaotong University + Tsinghua University; Jifeng Dai + Ping Luo + Lin Shao]

• HiBerNAC(arxiv2025.06) HiBerNAC: Hierarchical Brain-emulated Robotic Neural Agent Collective for Disentangling Complex Manipulation [arxiv link][Johns Hopkins Universit + Italian Institute of Technolog + University of Toronto + Harvard University]

• 👍SAPIEN(CVPR2020)(arxiv2020.03) SAPIEN: A SimulAted Part-based Interactive ENvironment [paper link][arxiv link][project link][code|official][UC San Diego + Stanford University + Simon Fraser University + Google Research + UC Los Angeles][SAPIEN is a realistic and physics-rich simulated environment that hosts a large-scale set for articulated objects. It enables various robotic vision and interaction tasks that require detailed part-level understanding. SAPIEN is a collaborative effort between researchers at UCSD, Stanford and SFU.]

• MPiNets(CoRL2023)(arxiv2022.10) Motion Policy Networks [openreview link][paper link][arxiv link][project link][code|official][University of Washington + NVIDIA; Dieter Fox]

• MimicGen(CoRL2023)(arxiv2023.10) MimicGen: A Data Generation System for Scalable Robot Learning using Human Demonstrations [openreview link][paper link][arxiv link][project link][code|official][NVIDIA + The University of Texas at Austin]

• 👍Gen2Sim(ICRA2024)(arxiv2023.10) Gen2Sim: Scaling up Robot Learning in Simulation with Generative Models [arxiv link][project link][code|official][CMU]

• 👍RoboGen(ICML2024)(arxiv2023.11) RoboGen: Towards Unleashing Infinite Data for Automated Robot Learning via Generative Simulation [arxiv link][project link][code|official][CMU + Tsinghua IIIS + MIT CSAIL + UMass Amherst + MIT-IBM AI Lab]

• GenSim(ICLR2024 spotlight)(arxiv2023.10) GenSim: Generating Robotic Simulation Tasks via Large Language Models [openreview link][arxiv link][project link][data link][code|official][MIT CSAIL + SJUT + UCSD + THU + UW + CMU; Xiaolong Wang]

• 3D-VLA(ICML2024)(arxiv2024.03) 3D-VLA: A 3D Vision-Language-Action Generative World Model [paper link][arxiv link][project link][UMA + SJTU + SCUT + WHU + MIT + UCLA]

• PhyRecon(arxiv2024.04) PhyRecon: Physically Plausible Neural Scene Reconstruction [arxiv link][project link][code|official][BIGAI + THU + PKU][It harnesses both differentiable rendering and differentiable physics simulation to achieve physically plausible scene reconstruction from multi-view images.]

• SAM-E(ICML2024)(arxiv2024.05) SAM-E: Leveraging Visual Foundation Model with Sequence Imitation for Embodied Manipulation [paper link][arxiv link][project link][weixin blog][code|official][THU + Shanghai AI Lab + HKUST][only tested on the dataset RLBench, and obtained inferior results than 3D Diffuser Actor]

• PhyScene(CVPR2024, Highlight)(arxiv2024.04) PhyScene: Physically Interactable 3D Scene Synthesis for Embodied AI [paper link][arxiv link][project link][code|official][BIGAI]

• SPIN(CVPR2024)(arxiv2024.05) SPIN: Simultaneous Perception, Interaction and Navigation [paper link][arxiv link][project link][CMU]

• IntervenGen(IROS2024)(arxiv2024.05) IntervenGen: Interventional Data Generation for Robust and Data-Efficient Robot Imitation Learning [paper link][arxiv link][project link][UC Berkeley + NVIDIA; Dieter Fox]

• 👍RoboCasa(RSS2024)(arxiv2024.06) RoboCasa: Large-Scale Simulation of Everyday Tasks for Generalist Robots [arxiv link][project link][weixin blog][zhihu blog][code|official][The University of Texas at Austin + NVIDIA Research; Real2Sim2Real]

• IRASim(arxiv2024.06) IRASim: Learning Interactive Real-Robot Action Simulators [arxiv link][project link][code|official][ByteDance Research + HKUST; Video Generation as Real-Robot Simulators]

• SimGen(arxiv2024.06) SimGen: Simulator-conditioned Driving Scene Generation [arxiv link][project link][code|official][University of California, Los Angeles + Shanghai Jiao Tong University; Minyi Guo]

• 👍Dreamitate(CoRL2024)(arxiv2024.06) Dreamitate: Real-World Visuomotor Policy Learning via Video Generation [openreview link][paper link][arxiv link][project link][code|official][Columbia University + Toyota Research Institute + Stanford University]

• GENIMA(arxiv2024.07) Generative Image as Action Models [arxiv link][project link][code|official][Dyson Robot Learning Lab; the last author is Stephen James][This is an interesting work with similar idea with Render and Diffuse]

• GRUtopia(arxiv2024.07) GRUtopia: Dream General Robots in a City at Scale [arxiv link][project link][code|official][OpenRobotLab, Shanghai AI Laboratory + Zhejiang University + Shanghai Jiao Tong University + Tsinghua University + Nanjing University + The Chinese University of Hong Kong + Xidian University]

• DiffusionForcing(arxiv2024.07) Diffusion Forcing: Next-token Prediction Meets Full-Sequence Diffusion [arxiv link][project link][code|official][MIT]

• RoboStudio(arxiv2024.08) RoboStudio: A Physics Consistent World Model for Robotic Arm with Hybrid Representation [arxiv link][project link][code|official (not released)][University of Southern California + National University of Singapore + University of Michigan + Peking University + The Hong Kong University of Science and Technology + Beijing Institute of Technology + Tsinghua University + Xiaomi Robot Technology + AiR, Tsinghua University]

• 👍👍Transfusion(arxiv2024.08) Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model [arxiv link][code|official][Meta + Waymo + University of Southern California]

• ICRT(arxiv2024.08) In-Context Imitation Learning via Next-Token Prediction [arxiv link][project link][code|official][UC Berkeley + Autodesk]

• 👍PhysGen(ECCV2024)(arxiv2024.09) PhysGen: Rigid-Body Physics-Grounded Image-to-Video Generation [arxiv link][project link][code|official][University of Illinois Urbana-Champaign]

• LINGO(SIGGRAPH2024)(arxiv2024.10) Autonomous Character-Scene Interaction Synthesis from Text Instruction [arxiv link][project link][Peking University + BIGAI][This paper introduces a framework for synthesizing multi-stage scene-aware interaction motions and a comprehensive language-annotated MoCap dataset (LINGO).]

• GenSim2(CoRL2024)(arxiv2024.10) GenSim2: Scaling Robot Data Generation with Multi-modal and Reasoning LLMs [openreview link][arxiv link][project link][code|official][Tsinghua University + UCSD + Shanghai Jiao Tong University + MIT CSAIL; Weinan Zhang + Huazhe Xu]

• SkillMimicGen(CoRL2024)(arxiv2024.10) SkillMimicGen: Automated Demonstration Generation for Efficient Skill Learning and Deployment [paper link][openreview link][arxiv link][project link][NVIDIA; Dieter Fox]

• DexMimicGen(ICRA2025)(arxiv2024.10) DexMimicGen: Automated Data Generation for Bimanual Dexterous Manipulation via Imitation Learning [arxiv link][project link][NVIDIA Research + UT Austin + UC San Diego; Yuke Zhu]

• RoboGSim(arxiv2024.11) RoboGSim: A Real2Sim2Real Robotic Gaussian Splatting Simulator [arxiv link][project link][Harbin Institute of Technology, Shenzhen + MEGVII Technology + Zhejiang University + Institute of Computing Technology, Chinese Academy of Sciences]

• DISCOVERSE(year2025) DISCOVERSE: Efficient Robot Simulation in Complex High-Fidelity Environments [paper link][project link][code|official][Tsinghua University + Zhejiang University + Tongji University + Xi'an Jiaotong University + D-Robotics]

• Cosmos-Predict1(arxiv2025.01) Cosmos World Foundation Model Platform for Physical AI [arxiv link][project link][code|official][NVIDIA]

• Cosmos-Transfer1(arxiv2025.03) Cosmos-Transfer1: Conditional World Generation with Adaptive Multimodal Control [arxiv link][project link][code|official][NVIDIA]

• Vid2World(arxiv2025.05) Vid2World: Crafting Video Diffusion Models to Interactive World Models [arxiv link][project link][Tsinghua University + Chongqing University; Mingsheng Long]

• FLARE(arxiv2025.05) FLARE: Robot Learning with Implicit World Modeling [arxiv link][project link][code|official][NVIDIA]

• FlowDreamer(arxiv2025.05) FlowDreamer: A RGB-D World Model with Flow-based Motion Representations for Robot Manipulation [arxiv link][project link][code|official][BIGAI + THU + BNU + HUST]

• 👍World4Omni(arxiv2025.06) World4Omni: A Zero-Shot Framework from Image Generation World Model to Robotic Manipulation [arxiv link][project link][National University of Singapore + NUS Guangzhou Research Translation and Innovation Institute + Shanghai Jiao Tong University + Tsinghua University + Peking University]

• Genie-Envisioner(arxiv2025.08) Genie Envisioner: A Unified World Foundation Platform for Robotic Manipulation [arxiv link][project link][code|official][AgiBot Genie Team + NUS LV-Lab + BUAA]

• BC-Transformer(robomimic)(CoRL2021 oral)(arxiv2021.08) What Matters in Learning from Offline Human Demonstrations for Robot Manipulation [openreview link][paper link][arxiv link][project link][code|official][Stanford University + The University of Texas at Austin][The proposed method BC-Transformer is used as a baseline in robocasa][robomimic: A Framework for Robot Learning from Demonstration. It offers a broad set of demonstration datasets collected on robot manipulation domains, and learning algorithms to learn from these datasets.]

• CLIPort(CoRL2021)(arxiv2021.09) CLIPort: What and Where Pathways for Robotic Manipulation [openreview link][paper link][arxiv link][project link][code|official][code|not official - CLIPort-Batchify][University of Washington + NVIDIA]

• BIP(IROS2022)(arxiv2022.08) A System for Imitation Learning of Contact-Rich Bimanual Manipulation Policies [paper link][arxiv link][project link][code|official][Carnegie Melon University + Intrinsic, An Alphabet Company + Arizona State University]

• BC-Z(CVPR2022)(arxiv2022.02) BC-Z: Zero-Shot Task Generalization with Robotic Imitation Learning [openreview link][paper link][arxiv link][project link][code|official][Robotics at Google + The Moonshot Factory + UC Berkeley + Stanford University; Chelsea Finn][It is based on the TensorFlow]

• ❤C2FARM(CVPR2022 Oral)(arxiv2021.06) Coarse-To-Fine Q-Attention: Efficient Learning for Visual Robotic Manipulation via Discretisation [paper link][arxiv link][project link][code|official][Dyson Robotics Lab, Imperial College London][It maybe the fisrt work to conduct next-best keyframe detection by the first author Stephen James, who also used this “predict the next (best) keyframe action” idea in his other works (RAL2022) Q-attention: Enabling Efficient Learning for Vision-based Robotic Manipulation and (TMLR2022) Auto-Lambda: Disentangling Dynamic Task Relationships. And the key-frames idea fisrtly proposed in work (ICRA2021) Coarse-to-Fine Imitation Learning: Robot Manipulation from a Single Demonstration by Edward Johns who leading the Robot Learning Lab at Imperial College London.]

• 👍CatBC(RSS2022)(arxiv2022.01) You Only Demonstrate Once: Category-Level Manipulation from Single Visual Demonstration [paper link][arxiv link][video link][Intrinsic Innovation LLC, CA, USA + Rutgers University]

• ❤R3M(CoRL2022)(arxiv2022.03) R3M: A Universal Visual Representation for Robot Manipulation [openreview link][paper link][arxiv link][project link][code|official][Stanford University + Meta AI; a pre-training method][We study if visual representations pre-trained on diverse human videos can enable efficient robotic manipulation. We pre-train a single representation, R3M, utilizing an objective that combines time contrastive learning, video-language alignment, and a sparsity penalty.]

• MVP(CoRL2022 oral)(arxiv2022.10) Real-World Robot Learning with Masked Visual Pre-training [openreview link][paper link][arxiv link][project link][code|official][University of California, Berkeley; a pre-training method][It first compiled a massive collection of 4.5 million images from ImageNet, Epic Kitchens, Something Something, 100 Days of Hands, and Ego4D datasets. Then, it pre-trained a model based on masked autoencoder (MAE).]

• ❤PerAct(CoRL2022)(arxiv2022.09) Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation [openreview link][paper link][arxiv link][project link][code|official][University of Washington + NVIDIA; Dieter Fox][It proposed a 3D policy that voxelizes the workspace and detects the next voxel action through global self-attention.][This work is largely based on C2FARM (CVPR2022) and PerceiverIO (ICLR2022); It constructs a structured observation and action space through keyframe extraction and voxelization following C2FARM.]

• ToolFlowNet(CoRL2022)(arxiv2022.11) ToolFlowNet: Robotic Manipulation with Tools via Predicting Tool Flow from Point Clouds [openreview link][paper link][arxiv link][project link][code|official][The Robotics Institute, Carnegie Mellon Universit]

• 👍CaP(ICRA2023)(arxiv2022.09) Code as Policies: Language Model Programs for Embodied Control [paper link][arxiv link][project link][code|official][Robotics at Google]

• 👍ProgPrompt(ICRA2023)(arxiv2022.09) ProgPrompt: Generating Situated Robot Task Plans using Large Language Models [paper link][arxiv link][project link][code|official][University of Southern California + NVIDIA; It has released code for replicating the results on the VirtualHome dataset.]

• PourIt(ICCV2023)(arxiv2023.07) PourIt!: Weakly-supervised Liquid Perception from a Single Image for Visual Closed-Loop Robotic Pouring [paper link][arxiv link][project link][code|official][Fudan University]

• Voltron(RSS2023)(arxiv2023.02) Language-Driven Representation Learning for Robotics [paper link][arxiv link][project link][code|official][Stanford University + Toyota Research Institute; a pre-training method][It provides code for loading pretrained Voltron, R3M, and MVP representations for adaptation to downstream tasks, as well as code for pretraining such representations on arbitrary datasets.]

• RoboNinja(RSS2023)(arxiv2023.02) RoboNinja: Learning an Adaptive Cutting Policy for Multi-Material Objects [paper link][arxiv link][project link][code|official][Columbia University + CMU + UC Berkeley + UC San Diego + UMass Amherst & MIT-IBM AI Lab]

• CherryBot(RSS2023)(arxiv2023.03) Cherry-Picking with Reinforcement Learning : Robust Dynamic Grasping in Unstable Conditions [paper link][arxiv link][project link][Carnegie Mellon University + University of Washington]

• MV-MWM(ICML2023)(arxiv2023.02) Multi-View Masked World Models for Visual Robotic Manipulation [paper link][arxiv link][project link][code|official][KAIST + Dyson Robot Learning Lab + Google Research + UC Berkeley; It used the less-popular TensorFlow 2]

• LLM-MCTS(NIPS2023)(arxiv2023.05) Large Language Models as Commonsense Knowledge for Large-Scale Task Planning [openreview link][paper link][arxiv link][project link][code|official][National University of Singapore][It used Large Language Models as both the commonsense world model and the heuristic policy within the Monte Carlo Tree Search framework, enabling better-reasoned decision-making for daily tasks.]

• L2M(NIPS2023)(arxiv2023.06) Learning to Modulate pre-trained Models in RL [openreview link][paper link][arxiv link][code|official][Johannes Kepler University Linz, Austria + Google DeepMind + UCL; tyr to adapt the already trained RL models.][Learning-to-Modulate]

• 👍RoboCLIP(NIPS2023)(arxiv2023.10) RoboCLIP: One Demonstration is Enough to Learn Robot Policies [paper link][arxiv link][project link][code|official][University of Southern California + University of California, Berkeley + Stanford University + Google Research; Chelsea Finn]

• MOO(CoRL2023)(arxiv2023.03) Open-World Object Manipulation using Pre-trained Vision-Language Models [openreview link][paper link][arxiv link][project link][Robotics at Google]

• HiveFormer(CoRL2023 Oral)(arxiv2022.09) Instruction-driven history-aware policies for robotic manipulations [openreview link][paper link][arxiv link][project link][code|official][Inria + IIIT Hyderabad; the second author Shizhe Chen][It is a 3D policy that enables attention between features of different history time steps.][It considered 74 tasks grouped into 9 categories on RLBench.]

• SGR(CoRL2023)(arxiv2023.06) A Universal Semantic-Geometric Representation for Robotic Manipulation [openreview link][paper link][arxiv link][project link][code|official][Tsinghua University + Shanghai Artificial Intelligence Laboratory + Shanghai Qi Zhi Institute; Yang Gao]

• BUDS/Stabilize2Act(CoRL2023)(arxiv2023.09) Stabilize to Act: Learning to Coordinate for Bimanual Manipulation [openreview link][paper link][arxiv link][project link][Stanford University]

• PolarNet(CoRL2023)(arxiv2023.09) PolarNet: 3D Point Clouds for Language-Guided Robotic [openreview link][paper link][arxiv link][project link][code|official][INRIA; the first author Shizhe Chen; 3D Vision-Language-Action][It is a 3D policy that computes dense point representations for the robot workspace using a PointNext backbone.][It considered 74 tasks grouped into 9 categories on RLBench following HiveFormer.]

• ❤RVT(CoRL2023 Oral)(arxiv2023.06) RVT: Robotic View Transformer for 3D Object Manipulation [openreview link][paper link][arxiv link][project link][code|official][NVIDIA; Dieter Fox][It re-projects the input RGB-D image to alternative image views, featurizes those and lifts the predictions to 3D to infer 3D locations for the robot’s end-effector.][It proposed a 3D policy that deploys a multi-view transformer to predict actions and fuses those across views by back-projecting to 3D.]

• ❤Act3D(CoRL2023)(arxiv2023.06) Act3D: 3D Feature Field Transformers for Multi-Task Robotic Manipulation [openreview link][paper link][arxiv link][project link][code|official][CMU; the first authors Theophile Gervet and Zhou Xian and Nikolaos Gkanatsios][It proposed a 3D policy that featurizes the robot’s 3D workspace using coarse-to-fine sampling and featurization.]

• GRIF(CoRL2023)(arxiv2023.07) Goal Representations for Instruction Following: A Semi-Supervised Language Interface to Control [openreview link][paper link][arxiv link][project link][code|official][University of California Berkeley + Microsoft Research; It used Semi-Supervised Learning and Contrastive Learning, but it also used the less-popular TensorFlow]

• GROOT(CoRL2023)(arxiv2023.10) Learning Generalizable Manipulation Policies with Object-Centric 3D Representations [openreview link][paper link][arxiv link][project link][code|official][The University of Texas, Austin + Sony AI; It used the SAM for segmenting out target objects.]

• 👍Optimus(CoRL2023)(arxiv2023.05) Imitating Task and Motion Planning with Visuomotor Transformers [openreview link][paper link][arxiv link][project link][code|official][CMU + NVIDIA; Dieter Fox][It is used as the baseline method by RoboCasa]

• 👍ShapeWarping(CoRL2023)(arxiv2023.06) One-shot Imitation Learning via Interation Warping [openreview link][paper link][arxiv link][project link][code|official][Northeastern University + Brown University + Microsoft Research + Google DeepMind + University of Amsterdam; Robert Platt]

• ScalingUp(CoRL2023)(arxiv2023.07) Scaling Up and Distilling Down: Language-Guided Robot Skill Acquisition [openreview link][paper link][arxiv link][project link][code|official][Columbia University + Google DeepMind][It used the Diffusion Policy for building a robust multi-task language-conditioned visuo-motor policy.]

• VoxPoser(CoRL2023 Oral)(arxiv2023.07) VoxPoser: Composable 3D Value Maps for Robotic Manipulation with Language Models [paper link][arxiv link][project link][code|official][by Stanford Fei-Fei Li; It extracts affordances and constraints from large language models (LLMs) and vision-language models (VLMs) to compose 3D value maps; It needs Detector+Segmentor+Tracker and thus is very slow]

• GNFactor(CoRL2023 Oral)(arxiv2023.08) GNFactor: Multi-Task Real Robot Learning with Generalizable Neural Feature Fields [openreview link][paper link][arxiv link][project link][code|official][SJTU + UC San Diego + University of Hong Kong + AWS AI, Amazon; a work by the Xiaolong Wang group][It proposed a 3D policy that co-optimizes a neural field for reconstructing the 3D voxels of the input scene and a PerAct module for predicting actions based on voxel representations.]

• Mechanisms(CoRL2023)(arxiv2023.11) Learning Reusable Manipulation Strategies [openreview link][paper link][arxiv link][project link][Massachusetts Institute of Technology]

• Diff-LfD(CoRL2023, Oral) Diff-LfD: Contact-aware Model-based Learning from Visual Demonstration for Robotic Manipulation via Differentiable Physics-based Simulation and Rendering [openreview link][paper link][project link][UC Berkeley + USTC + Zhejiang University + Nanjing University + University of Queensland + Shanghai Jiaotong University + National University of Singapore; Cewu Lu]

• 👍CPM(CoRL2023)(arxiv2024.05) Composable Part-Based Manipulation [paper link][openreview link][arxiv link][project link 1][project link 2][weixin blog][Stanford + MIT + NVIDIA + University of Utah + Georgia Tech + Jiajun Wu]