机器人 / 具身智能 - arXivDaily 专题

2403.00991 2024-10-08 cs.RO cs.CV cs.LG 83%

SELFI: Autonomous Self-Improvement with Reinforcement Learning for Social Navigation

Noriaki Hirose, Dhruv Shah, Kyle Stachowicz, Ajay Sridhar, Sergey Levine

专题命中模仿学习与强化学习：navigation(title)；robotic(abstract)；分类 cs.RO、cs.CV、cs.LG；robot learning(comments)

Comments 20pages, 12 figures, 2 tables, Conference on Robot Learning 2024

2401.03306 2024-01-09 cs.LG cs.AI cs.RO 83%

MOTO: Offline Pre-training to Online Fine-tuning for Model-based Robot Learning

Rafael Rafailov, Kyle Hatch, Victor Kolev, John D. Martin, Mariano Phielipp, Chelsea Finn

专题命中模仿学习与强化学习：robot learning(title,journal_ref)；manipulation(abstract)；分类 cs.RO、cs.AI、cs.LG

Comments This is an updated version of a manuscript that originally appeared at CoRL 2023. The project website is here https://sites.google.com/view/mo2o

Journal ref Proceedings of The 7th Conference on Robot Learning, PMLR 229:3654-3671, 2023

2210.02317 2023-06-28 cs.RO cs.AI cs.LG 83%

Real-Time Reinforcement Learning for Vision-Based Robotics Utilizing Local and Remote Computers

Yan Wang, Gautham Vasan, A. Rupam Mahmood

专题命中模仿学习与强化学习：robotics(title,comments)；robotic(abstract)；分类 cs.RO、cs.AI、cs.LG

Comments Appears in Proceedings of the 2023 International Conference on Robotics and Automation (ICRA). Source code at https://github.com/rlai-lab/relod and companion video at https://youtu.be/7iZKryi1xSY

2210.01969 2023-05-29 cs.LG cs.AI cs.RO 83%

Option-Aware Adversarial Inverse Reinforcement Learning for Robotic Control

Jiayu Chen, Tian Lan, Vaneet Aggarwal

专题命中模仿学习与强化学习：robotic(title,abstract)；分类 cs.RO、cs.AI、cs.LG；robotics(comments)

Comments This paper is partly presented at IEEE International Conference on Robotics and Automation (ICRA 2023)

2303.00085 2023-04-18 cs.RO cs.AI cs.LG 83%

AR3n: A Reinforcement Learning-based Assist-As-Needed Controller for Robotic Rehabilitation

Shrey Pareek, Harris Nisar, Thenkurussi Kesavadas

专题命中模仿学习与强化学习：robotic(title,abstract)；分类 cs.RO、cs.AI、cs.LG；robotics(journal_ref)

Comments 8 pages, 9 figures, IEEE RA-M

Journal ref IEEE Robotics and Automation Magazine, 2023

2209.07420 2023-02-10 cs.RO cs.AI cs.LG 83%

Scalable Task-Driven Robotic Swarm Control via Collision Avoidance and Learning Mean-Field Control

Kai Cui, Mengguang Li, Christian Fabian, Heinz Koeppl

专题命中模仿学习与强化学习：robotic(title,abstract)；分类 cs.RO、cs.AI、cs.LG；robotics(comments)

Comments Accepted to the 40th IEEE Conference on Robotics and Automation (ICRA)

2010.11940 2020-10-23 cs.RO cs.AI cs.LG 83%

Motion Planner Augmented Reinforcement Learning for Robot Manipulation in Obstructed Environments

Jun Yamada, Youngwoon Lee, Gautam Salhotra, Karl Pertsch, Max Pflueger, Gaurav S. Sukhatme, Joseph J. Lim, Peter Englert

专题命中模仿学习与强化学习：manipulation(title,abstract)；分类 cs.RO、cs.AI、cs.LG；robot learning(comments)

Comments Published at the Conference on Robot Learning (CoRL) 2020

1910.11670 2019-10-28 cs.RO cs.CV cs.LG 83%

Contextual Imagined Goals for Self-Supervised Robotic Learning

Ashvin Nair, Shikhar Bahl, Alexander Khazatsky, Vitchyr Pong, Glen Berseth, Sergey Levine

专题命中模仿学习与强化学习：robotic(title,abstract)；分类 cs.RO、cs.CV、cs.LG；robot learning(comments)

Comments 12 pages, to be presented at Conference on Robot Learning (CoRL) 2019. Project website: https://ccrig.github.io/

1511.03791 2015-11-16 cs.LG cs.CV cs.RO 83%

Towards Vision-Based Deep Reinforcement Learning for Robotic Motion Control

Fangyi Zhang, Jürgen Leitner, Michael Milford, Ben Upcroft, Peter Corke

专题命中模仿学习与强化学习：robotic(title,abstract)；分类 cs.RO、cs.CV、cs.LG；robotics(comments)

Comments 8 pages, to appear in the proceedings of Australasian Conference on Robotics and Automation (ACRA) 2015

2510.06277 2026-08-07 cs.CV cs.LG 版本更新 82%

Dynamic Object Masks as Goal Representations for Visual Goal-Conditioned Reinforcement Learning

动态目标掩码作为视觉目标条件强化学习的目标表示

Fahim Shahriar, Cheryl Wang, Alireza Azimi, Gautham Vasan, Hany Hamed, Abhishek Naik, A. Rupam Mahmood, Colin Bellinger

机构 * University of Alberta（阿尔伯塔大学）； McGill University（麦吉尔大学）； University of Ottawa（渥太华大学）； AMII ； CIFAR Canada AI Chair（CIFAR加拿大人工智能主席）； Vector Institute（向量研究所）

专题命中模仿学习与强化学习：robotics(abstract)；manipulation(abstract)；navigation(abstract)；robotic(abstract)

AI总结该研究针对视觉目标条件强化学习，提出动态目标掩码作为目标表示，结合Detic等预训练检测器生成掩码，在机械臂和仿真导航任务中实现高成功率与高效学习，支持仿真到现实迁移。

2603.01452 2026-03-03 cs.AI cs.RO 82%

Scaling Tasks, Not Samples: Mastering Humanoid Control through Multi-Task Model-Based Reinforcement Learning

通过多任务模型基于强化学习掌握人形控制

Shaohuai Liu, Weirui Ye, Yilun Du, Le Xie

专题命中模仿学习与强化学习：robotics(abstract)；embodied AI(abstract)；world model(abstract)；robotic(abstract)

AI总结本文提出EfficientZero-Multitask算法，通过多任务模型基于强化学习提升人形机器人控制性能，实现高效样本利用和高任务适应性。

2304.08743 2023-06-30 cs.LG cs.RO 82%

Benchmarking Actor-Critic Deep Reinforcement Learning Algorithms for Robotics Control with Action Constraints

Kazumi Kasaura, Shuwa Miura, Tadashi Kozuno, Ryo Yonetani, Kenta Hoshino, Yohei Hosoe

专题命中模仿学习与强化学习：robotics(title,abstract)；分类 cs.RO、cs.LG

Comments 8 pages, 7 figures, accepted to Robotics and Automation Letters

Journal ref IEEE Robotics and Automation Letters 8(8) (2023) 4449-4456

2306.10985 2023-06-21 cs.CL cs.LG cs.RO 82%

LARG, Language-based Automatic Reward and Goal Generation

Julien Perez, Denys Proux, Claude Roux, Michael Niemaz

专题命中模仿学习与强化学习：robot learning(abstract)；manipulation(abstract)；navigation(abstract)；robotic(abstract)

2204.07404 2023-04-14 cs.AI cs.RO 82%

Divide & Conquer Imitation Learning

Alexandre Chenu, Nicolas Perrin-Gilbert, Olivier Sigaud

专题命中模仿学习与强化学习：robotics(abstract)；manipulation(abstract)；navigation(abstract)；robotic(abstract)

2010.08252 2021-03-26 cs.RO cs.AI 82%

Hyperparameter Auto-tuning in Self-Supervised Robotic Learning

Jiancong Huang, Juan Rojas, Matthieu Zimmer, Hongmin Wu, Yisheng Guan, Paul Weng

专题命中模仿学习与强化学习：robotic(title)；robot learning(abstract)；robotics(comments,journal_ref)；分类 cs.RO、cs.AI

Comments 8 pages, 6 figures, Published in IEEE Robotics and Automation Letters; Presented at The 2021 International Conference on Robotics and Automation (ICRA 2021); Presented at Deep RL Workshop, NeurIPS 2020

Journal ref IEEE Robotics and Automation Letters, Volume:6, Issue:2, P. 3537-3544, April 2021

1811.06187 2018-11-16 cs.RO cs.AI 82%

Intervention Aided Reinforcement Learning for Safe and Practical Policy Optimization in Navigation

Fan Wang, Bo Zhou, Ke Chen, Tingxiang Fan, Xi Zhang, Jiangyong Li, Hao Tian, Jia Pan

专题命中模仿学习与强化学习：navigation(title,abstract)；分类 cs.RO、cs.AI；robot learning(journal_ref)

Journal ref Wang, F., Zhou, B., Chen, K., Fan, T., Zhang, X., Li, J., ... & Pan, J. (2018, October). Intervention Aided Reinforcement Learning for Safe and Practical Policy Optimization in Navigation. In Conference on Robot Learning (pp. 410-421)

2603.13707 2026-07-31 cs.RO cs.AI cs.LG 版本更新 82%

REFINE-DP: Diffusion Policy Fine-tuning for Humanoid Loco-manipulation via Reinforcement Learning

REFINE-DP：通过强化学习实现人形机器人的动态-操作协调政策微调

Zhaoyuan Gu, Yipu Chen, Zimeng Chai, Alfred Cueva, Thong Nguyen, Yifan Wu, Huishu Xue, Minji Kim, Isaac Legene, Fukang Liu, KyoungMok Kim, Ayan Barula, Yongxin Chen, Ye Zhao

机构 * The Institute for Robotics and Intelligent Machines（机器人与智能机械研究所）

专题命中模仿学习与强化学习：manipulation(title,abstract)；分类 cs.RO、cs.AI、cs.LG

AI总结本文提出REFINE-DP框架，通过联合优化扩散政策高层规划器和基于强化学习的低层动态-操作控制器，提升人形机器人在复杂环境中的任务成功率与执行稳定性。

2507.21638 2026-07-22 cs.AI cs.LG cs.MA cs.RO 版本更新 82%

Assistax: A Multi-Agent Hardware-Accelerated Reinforcement Learning Benchmark for Assistive Robotics

Assistax: 一个用于辅助机器人的多智能体硬件加速强化学习基准

Leonard Hinckeldey, Elliot Fosong, Rimvydas Rubavicius, Elle Miller, Trevor McInroe, Fan Zhang, Patricia Wollstadt, Stefano V. Albrecht, Subramanian Ramamoorthy

机构 * University of California, Berkeley（加州大学伯克利分校）； Stanford University（斯坦福大学）

专题命中模仿学习与强化学习：robotics(title,abstract)；分类 cs.RO、cs.AI、cs.LG

AI总结提出Assistax基准，利用JAX硬件加速和基于多智能体强化学习的辅助机器人任务，实现高达370倍加速，并测试机器人的零样本协调能力。

Comments Accepted at the Reinforcement Learning Conference 2026

2607.10991 2026-07-14 cs.RO cs.AI cs.CV 新提交 82%

Think When It Matters: Conditional VLM Reasoning for Social Navigation with RL Policies

在重要时刻思考：用于社交导航的基于RL策略的条件VLM推理

Ali Ahmadi, Hamed Rahimi, Adrien Jacquet Cretides, Marie Samson, Mahdi Khoramshahi, Mohamed Chetouani

机构 * Institut des Systèmes Intelligents et de Robotique (ISIR)（智能系统与机器人研究所）

专题命中模仿学习与强化学习：navigation(title,abstract)；分类 cs.RO、cs.AI、cs.CV

AI总结研究社交机器人导航中强化学习策略缺乏语义推理能力的问题，提出HUMA混合架构，动态平衡RL策略与VLM的优势。在基准测试中任务成功率提高，减少碰撞，消融研究验证组件，实际部署证明方法可行。

Comments CoRL 2026 submission. 15 pages, 7 figures

2605.02867 2026-06-23 cs.LG cs.AI cs.RO 版本更新 82%

Enhancing RL Generalizability in Robotics through SHAP Analysis of Algorithms and Hyperparameters

通过SHAP分析算法和超参数增强机器人学中强化学习的泛化能力

Lingxiao Kong, Cong Yang, Oya Deniz Beyan, Zeyd Boukhers

机构 * Fraunhofer Institute for Applied Information Technology（弗劳恩霍夫应用信息科技研究所）； University of Cologne（科隆大学）； University Hospital of Cologne（科隆大学医院）

专题命中模仿学习与强化学习：robotics(title)；robotic(abstract)；分类 cs.RO、cs.AI、cs.LG

AI总结提出基于SHAP的可解释框架，量化算法和超参数对强化学习泛化差距的贡献，并用于配置选择以提升泛化能力。

Comments 16 pages, 7 figures, accepted by ICPR 2026

2509.26633 2026-06-17 cs.RO cs.AI cs.LG cs.SY eess.SY 版本更新 82%

OmniRetarget: Interaction-Preserving Data Generation for Humanoid Whole-Body Loco-Manipulation and Scene Interaction

OmniRetarget：面向人形全身运动操控与场景交互的交互保持数据生成

Lujie Yang, Xiaoyu Huang, Zhen Wu, Angjoo Kanazawa, Pieter Abbeel, Carmelo Sferrazza, C. Karen Liu, Rocky Duan, Guanya Shi

机构 * Amazon FAR (Frontier AI & Robotics)（亚马逊前沿人工智能与机器人实验室）； MIT（麻省理工学院）； UC Berkeley（伯克利大学）； Stanford University（斯坦福大学）； CMU（卡内基梅隆大学）

专题命中模仿学习与强化学习：manipulation(title,abstract)；分类 cs.RO、cs.AI、cs.LG

AI总结提出OmniRetarget引擎，通过交互网格显式建模并保持智能体、地形和物体间的空间与接触关系，将人类运动重定向为机器人运动，生成高质量轨迹以训练强化学习策略，实现长时间跑酷和操控技能。

Comments Project website: https://omniretarget.github.io

2207.09845 2026-04-29 cs.RO cs.AI cs.HC cs.LG 82%

Quantifying the Effect of Feedback Frequency in Interactive Reinforcement Learning for Robotic Tasks

量化交互强化学习在机器人任务中的反馈频率效应

Daniel Harnack, Julie Pivin-Bachler, Nicolás Navarro-Guerrero

机构 * Robotics and Interactive Systems – UPSSITECH, University Paul Sabatier（机器人与交互系统——UPSSITECH，保罗·萨巴蒂埃大学）

专题命中模仿学习与强化学习：robotic(title,abstract)；分类 cs.RO、cs.AI、cs.LG

AI总结研究探讨了反馈频率对机器人任务中交互强化学习效率的影响，通过不同复杂度的机械臂逆运动学学习实验，发现反馈频率需随任务熟练度动态调整，无单一最优频率。

Comments Neural Computing and Applications (2022). Special Issue on Human-aligned Reinforcement Learning for Autonomous Agents and Robots

2601.06133 2026-02-10 cs.LG cs.AI cs.RO 82%

A Review of Online Diffusion Policy RL Algorithms for Scalable Robotic Control

在线扩散策略强化学习算法综述：可扩展机器人控制

Wonhyeok Choi, Shutong Ding, Minwoo Choi, Jungwan Woo, Kyumin Hwang, Jaeyeul Kim, Ye Shi, Sunghoon Im

机构 * Daegu Gyeongbuk Institute of Science and Technology（大邱庆尚科学技术院）； ShanghaiTech University（上海科技大学）

专题命中模仿学习与强化学习：robotic(title,abstract)；分类 cs.RO、cs.AI、cs.LG

AI总结本文综述了在线扩散策略强化学习算法，分析了其在可扩展机器人控制中的性能、挑战及未来发展方向。

2505.18417 2026-02-02 cs.RO cs.AI cs.LG 82%

Reinforcement Learning for Ballbot Navigation in Uneven Terrain

基于强化学习的Ballbot在不平整地形中的导航

Achkan Salehi

机构 * VimaLabs（维马实验室）

专题命中模仿学习与强化学习：navigation(title,abstract)；分类 cs.RO、cs.AI、cs.LG

AI总结本文提出基于MuJoCo的开源Ballbot模拟器，展示了通过强化学习方法有效导航不平整地形的成果。

Comments 6 pages, 9 figures, 2 tables. Version two corrects figure 4 and adds some experiments

2601.14140 2026-01-21 cs.AR 82%

CREATE: Cross-Layer Resilience Characterization and Optimization for Efficient yet Reliable Embodied AI Systems

CREATE: 跨层韧性表征与优化以实现高效且可靠的具身AI系统

Tong Xie, Yijiahao Qi, Jinqi Wen, Zishen Wan, Yanchi Dong, Zihao Wang, Shaofei Cai, Yitao Liang, Tianyu Jia, Yuan Wang, Runsheng Wang, Meng Li

专题命中模仿学习与强化学习：embodied AI(title,abstract)；embodied agent(abstract)

AI总结 CREATE提出了一种跨层韧性优化方法，通过电路、模型和应用层面的协同设计，实现高效且可靠的具身AI系统。

Comments 18 pages, 21 figures. Accepted by ASPLOS 2026

2601.07821 2026-01-13 cs.RO cs.AI cs.LG 82%

Failure-Aware RL: Reliable Offline-to-Online Reinforcement Learning with Self-Recovery for Real-World Manipulation

具有自恢复能力的失败感知强化学习：用于现实世界操控的可靠离线到在线强化学习

Huanyu Li, Kun Lei, Sheng Zang, Kaizhe Hu, Yongyuan Liang, Bo An, Xiaoli Li, Huazhe Xu

机构 * Shanghai Qi Zhi Institute（上海启智研究院）； Shanghai Jiao Tong University（上海交通大学）； IIIS, Tsinghua University（清华大学人工智能研究院）； Nanyang Technological University（南洋理工大学）； A*STAR Institute for Infocomm Research（新加坡科技动力研究院）； University of Maryland（马里兰大学）

专题命中模仿学习与强化学习：manipulation(title)；robotic(abstract)；分类 cs.RO、cs.AI、cs.LG

AI总结本研究提出FARL框架，通过整合安全批评者和恢复策略，有效减少现实世界强化学习中的失败并提升性能。

Comments Project page: https://failure-aware-rl.github.io

2505.16394 2025-10-28 cs.RO cs.AI cs.CV 82%

Raw2Drive: Reinforcement Learning with Aligned World Models for End-to-End Autonomous Driving (in CARLA v2)

Zhenjie Yang, Xiaosong Jia, Qifeng Li, Xue Yang, Maoqing Yao, Junchi Yan

机构 * Sch. of CS, Sch. of AIS, Sch. of AI, Shanghai Jiao Tong University（计算机科学学院、人工智能科学学院、人工智能学院，上海交通大学）； Institute of Trustworthy Embodied AI, Fudan University（可信具身人工智能研究院，复旦大学）； AgiBot

专题命中模仿学习与强化学习：world model(title,abstract)；分类 cs.RO、cs.AI、cs.CV

Comments Accepted by NeurIPS 2025

2408.14183 2025-09-30 cs.RO cs.AI cs.LG 82%

Robot Navigation with Entity-Based Collision Avoidance using Deep Reinforcement Learning

Yury Kolomeytsev, Dmitry Golembiovsky

机构 * Lomonosov Moscow State University, Moscow, Russia（罗蒙诺索夫莫斯科国立大学，莫斯科，俄罗斯）

专题命中模仿学习与强化学习：navigation(title,abstract)；分类 cs.RO、cs.AI、cs.LG

Comments 15 pages, 4 figures

2508.06571 2025-08-18 cs.AI cs.CV cs.RO 82%

IRL-VLA: Training an Vision-Language-Action Policy via Reward World Model

Anqing Jiang, Yu Gao, Yiru Wang, Zhigang Sun, Shuo Wang, Yuwen Heng, Hao Sun, Shichen Tang, Lijuan Zhu, Jinhao Chai, Jijun Wang, Zichong Gu, Hao Jiang, Li Sun

机构 * Tsinghua University（清华大学）

专题命中模仿学习与强化学习：world model(title,abstract)；分类 cs.RO、cs.AI、cs.CV

Comments 9 pagres, 2 figures

2501.18016 2025-06-11 cs.RO cs.AI cs.LG cs.SY eess.SY 82%

Digital Twin Synchronization: Bridging the Sim-RL Agent to a Real-Time Robotic Additive Manufacturing Control

Matsive Ali, Sandesh Giri, Sen Liu, Qin Yang

机构 * Department of Mechanical Engineering, University of Louisiana at Lafayette（路易斯安那大学拉斐特分校机械工程系）； Computer Science & Information Systems Department, Bradley University（布雷纳德大学计算机科学与信息系统系）

专题命中模仿学习与强化学习：robotic(title)；robotics(abstract)；分类 cs.RO、cs.AI、cs.LG

Comments This paper had been accepted by the 2025 IEEE International Conference on Engineering Reliable Autonomous Systems (ERAS)

1. 模仿学习与强化学习 4106 篇

SELFI: Autonomous Self-Improvement with Reinforcement Learning for Social Navigation

MOTO: Offline Pre-training to Online Fine-tuning for Model-based Robot Learning

Real-Time Reinforcement Learning for Vision-Based Robotics Utilizing Local and Remote Computers

Option-Aware Adversarial Inverse Reinforcement Learning for Robotic Control

AR3n: A Reinforcement Learning-based Assist-As-Needed Controller for Robotic Rehabilitation

Scalable Task-Driven Robotic Swarm Control via Collision Avoidance and Learning Mean-Field Control

Motion Planner Augmented Reinforcement Learning for Robot Manipulation in Obstructed Environments

Contextual Imagined Goals for Self-Supervised Robotic Learning

Towards Vision-Based Deep Reinforcement Learning for Robotic Motion Control

Dynamic Object Masks as Goal Representations for Visual Goal-Conditioned Reinforcement Learning

Scaling Tasks, Not Samples: Mastering Humanoid Control through Multi-Task Model-Based Reinforcement Learning

Benchmarking Actor-Critic Deep Reinforcement Learning Algorithms for Robotics Control with Action Constraints

LARG, Language-based Automatic Reward and Goal Generation

Divide & Conquer Imitation Learning

Hyperparameter Auto-tuning in Self-Supervised Robotic Learning

Intervention Aided Reinforcement Learning for Safe and Practical Policy Optimization in Navigation

REFINE-DP: Diffusion Policy Fine-tuning for Humanoid Loco-manipulation via Reinforcement Learning

Assistax: A Multi-Agent Hardware-Accelerated Reinforcement Learning Benchmark for Assistive Robotics

Think When It Matters: Conditional VLM Reasoning for Social Navigation with RL Policies

Enhancing RL Generalizability in Robotics through SHAP Analysis of Algorithms and Hyperparameters

OmniRetarget: Interaction-Preserving Data Generation for Humanoid Whole-Body Loco-Manipulation and Scene Interaction

Quantifying the Effect of Feedback Frequency in Interactive Reinforcement Learning for Robotic Tasks

A Review of Online Diffusion Policy RL Algorithms for Scalable Robotic Control

Reinforcement Learning for Ballbot Navigation in Uneven Terrain

CREATE: Cross-Layer Resilience Characterization and Optimization for Efficient yet Reliable Embodied AI Systems

Failure-Aware RL: Reliable Offline-to-Online Reinforcement Learning with Self-Recovery for Real-World Manipulation

Raw2Drive: Reinforcement Learning with Aligned World Models for End-to-End Autonomous Driving (in CARLA v2)

Robot Navigation with Entity-Based Collision Avoidance using Deep Reinforcement Learning

IRL-VLA: Training an Vision-Language-Action Policy via Reward World Model

Digital Twin Synchronization: Bridging the Sim-RL Agent to a Real-Time Robotic Additive Manufacturing Control