机器人 / 具身智能 - arXivDaily 专题

2512.00062 2026-06-02 cs.RO cs.AI cs.LG 75%

SpeedAug: Policy Acceleration via Tempo-Enriched Policy and RL Fine-Tuning

SpeedAug: 通过节奏增强策略和强化学习微调实现策略加速

Taewook Nam, Junmo Cho, Youngsoo Jang, Sung Ju Hwang

机构 * KAIST（韩国科学技术院）； UNIST（全南大学）； DeepAuto.ai

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

AI总结提出SpeedAug框架，通过节奏增强先验策略和强化学习微调，使机器人策略学习任务最优执行节奏，在保持高成功率的同时显著提升执行速度和样本效率。

2605.19503 2026-05-21 cs.RO cs.AI cs.LG 75%

ARC-RL: A Reinforcement Learning Playground Inspired by ARC Raiders

ARC-RL: 一种受ARC Raiders启发的强化学习游乐场

Carlo Romeo, Andrew D. Bagdanov

机构 * Media Integration and Communication Center – University of Florence（媒体整合与通信中心——佛罗伦萨大学）

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

AI总结本文提出ARC-RL，一个包含四种MuJoCo连续控制环境的强化学习游乐场，这些环境的机器人形态灵感来自ARC Raiders的生物目录，通过统一的观察模板、动作约定和奖励函数，研究不同形态和动画风格约束下的强化学习算法性能。

2505.21410 2026-04-22 cs.AI cs.LG cs.RO 75%

MRS: Multi-Resolution Skills for HRL Agents

MRS：多分辨率技能用于HRL智能体

Shashank Sharma, Janina Hoffmann, Vinay Namboodiri

机构 * DeepMind

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

AI总结本文提出MRS多分辨率技能，通过学习多个固定时间跨度的目标预测模块，解决HRL中子目标选择精度与预测噪声的矛盾，提升长周期任务性能。

2604.13504 2026-04-16 cs.LG cs.AI cs.CL cs.MA cs.RO 75%

Chain of Uncertain Rewards with Large Language Models for Reinforcement Learning

基于大语言模型的强化学习中不确定奖励链

Shentong Mo

机构 * Carnegie Mellon University（卡内基梅隆大学）

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

AI总结本文提出CoUR框架，利用大语言模型优化强化学习中的奖励函数设计，通过不确定性量化和贝叶斯优化提升效率与鲁棒性。

2604.00557 2026-04-02 cs.RO cs.CV cs.LG 75%

Multi-Camera View Scaling for Data-Efficient Robot Imitation Learning

多摄像头视角缩放用于数据高效的机器人模仿学习

Yichen Xie, Yixiao Wang, Shuqi Zhao, Cheng-En Wu, Masayoshi Tomizuka, Jianwen Xie, Hao-Shu Fang

机构 * UC Berkeley（加州大学伯克利分校）； Lambda, Inc（Lambda公司）； MIT（麻省理工学院）

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

AI总结本文提出一种利用场景多样性提升机器人模仿学习数据效率的方法，通过摄像头视角缩放生成伪演示，增强视觉表示的视角不变性，并在仿真和现实任务中验证了其有效性。

2510.07151 2026-03-05 cs.LG cs.AI cs.RO 75%

ELMUR: External Layer Memory with Update/Rewrite for Long-Horizon RL Problems

ELMUR：带有更新/重写的外部层记忆用于长周期强化学习问题

Egor Cherepanov, Alexey K. Kovalev, Aleksandr I. Panov

机构 * AXXX ； MIRIAI

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

AI总结 ELMUR通过结构化外部记忆提升长周期强化学习任务中的决策能力，实现显著性能提升。

Comments 31 pages, 15 figures, 8 tables

2602.19372 2026-02-24 cs.RO cs.CV cs.LG 75%

Seeing Farther and Smarter: Value-Guided Multi-Path Reflection for VLM Policy Optimization

看得更远且更聪明：基于价值引导的多路径反射用于VLM策略优化

Yanting Yang, Shenyuan Gao, Qingwen Bu, Li Chen, Dimitris N. Metaxas

机构 * Rutgers University（罗格斯大学）； The Hong Kong University of Science and Technology（香港科学与技术大学）； The University of Hong Kong（香港大学）

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

AI总结本研究提出了一种基于价值引导的多路径反射方法，用于提升VLM在机器人操作任务中的策略优化性能，通过解耦状态评估与动作生成，提高决策鲁棒性并减少推理时间。

Comments ICRA 2026

2602.14526 2026-02-17 cs.RO cs.AI cs.LG 75%

TWISTED-RL: Hierarchical Skilled Agents for Knot-Tying without Human Demonstrations

TWISTED-RL：无人类示范的分层技能代理用于打结

Guy Freund, Tom Jurgenson, Matan Sudry, Erez Karpas

机构 * Reichman University（雷赫曼大学）； Technion – Israel Institute of Technology（技术学院——以色列理工学院）； The University of Texas at Austin（德克萨斯大学奥斯汀分校）

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

AI总结 TWISTED-RL通过强化学习和拓扑动作实现无示范的复杂打结任务，提升泛化能力和效率。

2503.05696 2026-02-13 cs.LG cs.AI cs.RO 75%

A Multi-Fidelity Control Variate Approach for Policy Gradient Estimation

一种多保真度控制变量化法用于策略梯度估计

Xinjie Liu, Cyrus Neary, Kushagra Gupta, Wesley A. Suttle, Christian Ellis, Ufuk Topcu, David Fridovich-Keil

机构 * The University of Texas at Austin（德克萨斯大学奥斯汀分校）； The University of British Columbia（不列颠哥伦比亚大学）； DEVCOM Army Research Laboratory（国防部陆军研究实验室）

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

AI总结本文提出一种多保真度控制变量化方法，通过结合稀缺目标环境数据和丰富的低保真度模拟数据，提高策略梯度估计的样本效率和收敛速度，适用于有限高保真度数据但丰富低保真度数据的机器人任务。

2506.09049 2026-01-23 cs.AI cs.CV cs.RO 75%

VIKI-R: Coordinating Embodied Multi-Agent Cooperation via Reinforcement Learning

VIKI-R: 通过强化学习协调具身多智能体合作

Li Kang, Xiufeng Song, Heng Zhou, Yiran Qin, Jie Yang, Xiaohong Liu, Philip Torr, Lei Bai, Zhenfei Yin

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

AI总结 VIKI-R通过强化学习协调多智能体合作，提出分层基准VIKI-Bench，显著提升多智能体视觉驱动合作性能。

Comments Accepted by NeurIPS 2025 Track on Datasets and Benchmarks. Project page: https://faceong.github.io/VIKI-R/

2512.03911 2025-12-04 cs.RO cs.AI cs.LG 75%

Autonomous Reinforcement Learning Robot Control with Intel's Loihi 2 Neuromorphic Hardware

自主强化学习机器人控制与英特尔Loihi 2神经形态硬件

Kenneth Stewart, Roxana Leontie, Samantha Chapin, Joe Hays, Sumit Bam Shrestha, Carl Glen Henshaw

机构 * U.S. Naval Research Laboratory Naval（美国海军研究实验室）； Intel Labs（英特尔实验室）

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

AI总结本文提出将强化学习策略转换为神经形态硬件的端到端流程，用于实现低功耗的机器人控制。

Comments Submitted for review at NICE 2026 (Neuro-Inspired Computational Elements) conference

2411.12155 2025-11-18 cs.LG cs.AI cs.RO 75%

Coarse-to-fine Q-Network with Action Sequence for Data-Efficient Reinforcement Learning

Younggyo Seo, Pieter Abbeel

机构 * UC Berkeley（伯克利大学）

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

Comments 18 Pages. Website: https://younggyo.me/cqn-as/

2510.19495 2025-10-28 cs.RO cs.AI cs.LG 75%

Using Non-Expert Data to Robustify Imitation Learning via Offline Reinforcement Learning

Kevin Huang, Rosario Scalise, Cleah Winston, Ayush Agrawal, Yunchu Zhang, Rohan Baijal, Markus Grotz, Byron Boots, Benjamin Burchfiel, Masha Itkina, Paarth Shah, Abhishek Gupta

机构 * University of Washington（华盛顿大学）； Toyota Research Institute (TRI)（丰田研究院）

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

2509.09674 2025-09-12 cs.RO cs.AI cs.CL cs.LG 75%

SimpleVLA-RL: Scaling VLA Training via Reinforcement Learning

Haozhan Li, Yuxin Zuo, Jiale Yu, Yuhao Zhang, Zhaohui Yang, Kaiyan Zhang, Xuekai Zhu, Yuchen Zhang, Tianxing Chen, Ganqu Cui, Dehui Wang, Dingxiang Luo, Yuchen Fan, Youbang Sun, Jia Zeng, Jiangmiao Pang, Shanghang Zhang, Yu Wang, Yao Mu, Bowen Zhou, Ning Ding

机构 * Shanghai Jiao Tong University（上海交通大学）； Peking University（北京大学）； The University of Hong Kong（香港大学）； Shanghai AI Lab（上海人工智能实验室）

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

2508.06767 2025-08-12 cs.LG cs.AI cs.DC cs.MA cs.RO 75%

PANAMA: A Network-Aware MARL Framework for Multi-Agent Path Finding in Digital Twin Ecosystems

Arman Dogru, R. Irem Bor-Yaliniz, Nimal Gamini Senarath

机构 * Huawei Canada Advanced Research Center（华为加拿大高级研究中心）

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

2507.19151 2025-08-05 cs.RO cs.AI cs.LG cs.MA 75%

ReCoDe: Reinforcement Learning-based Dynamic Constraint Design for Multi-Agent Coordination

Michael Amir, Guang Yang, Zhan Gao, Keisuke Okumura, Heedo Woo, Amanda Prorok

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

Comments To appear in CoRL 2025

2506.07744 2025-07-08 cs.LG cs.AI cs.RO 75%

Graph-Assisted Stitching for Offline Hierarchical Reinforcement Learning

Seungho Baek, Taegeon Park, Jongchan Park, Seungjun Oh, Yusung Kim

机构 * Department of Computer Science and Engineering, Sungkyunkwan University, Suwon, Republic of Korea（计算机科学与工程系，成均馆大学，苏旺，韩国）； Department of Artificial Intelligence, Sungkyunkwan University, Suwon, Republic of Korea（人工智能系，成均馆大学，苏旺，韩国）

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

Comments ICML 2025

2402.05421 2025-06-16 cs.LG cs.AI cs.RO 75%

DiffTORI: Differentiable Trajectory Optimization for Deep Reinforcement and Imitation Learning

Weikang Wan, Ziyu Wang, Yufei Wang, Zackory Erickson, David Held

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

Comments NeurIPS 2024 (Spotlight)

2407.21244 2025-05-23 cs.RO cs.AI cs.CV 75%

VITAL: Interactive Few-Shot Imitation Learning via Visual Human-in-the-Loop Corrections

Hamidreza Kasaei, Mohammadreza Kasaei

机构 * Department of Artificial Intelligence, University of Groningen（格罗宁根大学人工智能系）； School of Informatics, University of Edinburgh（爱丁堡大学信息学院）

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

2505.02833 2025-05-06 cs.RO cs.CV cs.LG 75%

TWIST: Teleoperated Whole-Body Imitation System

Yanjie Ze, Zixuan Chen, João Pedro Araújo, Zi-ang Cao, Xue Bin Peng, Jiajun Wu, C. Karen Liu

机构 * Stanford University（斯坦福大学）； Simon Fraser University（西蒙弗雷泽大学）

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

Comments Project website: https://humanoid-teleop.github.io

2310.17178 2025-03-21 cs.AI cs.LG cs.RO 75%

Relational Object-Centric Actor-Critic

Leonid Ugadiarov, Vitaliy Vorobyov, Aleksandr I. Panov

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

2503.09985 2025-03-20 cs.RO cs.CV cs.LG 75%

ES-Parkour: Advanced Robot Parkour with Bio-inspired Event Camera and Spiking Neural Network

Qiang Zhang, Jiahang Cao, Jingkai Sun, Yecheng Shao, Gang Han, Wen Zhao, Yijie Guo, Renjing Xu

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

2503.01871 2025-03-05 cs.LG cs.AI cs.RO 75%

Data Augmentation for Instruction Following Policies via Trajectory Segmentation

Niklas Höpner, Ilaria Tiddi, Herke van Hoof

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

2502.05526 2025-02-11 cs.RO cs.AI cs.LG cs.MA 75%

Towards Learning Scalable Agile Dynamic Motion Planning for Robosoccer Teams with Policy Optimization

Brandon Ho, Batuhan Altundas, Matthew Gombolay

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

2501.16664 2025-01-29 cs.RO cs.CV cs.LG 75%

Improving Vision-Language-Action Model with Online Reinforcement Learning

Yanjiang Guo, Jianke Zhang, Xiaoyu Chen, Xiang Ji, Yen-Jen Wang, Yucheng Hu, Jianyu Chen

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

Comments Accepted to ICRA 2025

2412.10713 2024-12-17 cs.LG cs.AI cs.CR cs.RO 75%

RAT: Adversarial Attacks on Deep Reinforcement Agents for Targeted Behaviors

Fengshuo Bai, Runze Liu, Yali Du, Ying Wen, Yaodong Yang

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

Comments Accepted by AAAI 2025

2405.16194 2024-11-27 cs.LG cs.AI cs.RO 75%

Diffusion-Reward Adversarial Imitation Learning

Chun-Mao Lai, Hsiang-Chun Wang, Ping-Chun Hsieh, Yu-Chiang Frank Wang, Min-Hung Chen, Shao-Hua Sun

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

Comments NeurIPS 2024. Project page: https://nturobotlearninglab.github.io/DRAIL/

2411.01396 2024-11-05 cs.LG cs.AI cs.RO 75%

Exploring the Edges of Latent State Clusters for Goal-Conditioned Reinforcement Learning

Yuanlin Duan, Guofeng Cui, He Zhu

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

Comments NeurIPS2024 Poster

2402.17257 2024-10-29 cs.LG cs.AI cs.RO 75%

RIME: Robust Preference-based Reinforcement Learning with Noisy Preferences

Jie Cheng, Gang Xiong, Xingyuan Dai, Qinghai Miao, Yisheng Lv, Fei-Yue Wang

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

Comments Accepted by ICML 2024 (Spotlight, top 3.5%)

2409.16578 2024-10-02 cs.RO cs.CV cs.LG 75%

FLaRe: Achieving Masterful and Adaptive Robot Policies with Large-Scale Reinforcement Learning Fine-Tuning

Jiaheng Hu, Rose Hendrix, Ali Farhadi, Aniruddha Kembhavi, Roberto Martin-Martin, Peter Stone, Kuo-Hao Zeng, Kiana Ehsani

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

1. 模仿学习与强化学习 4111 篇

SpeedAug: Policy Acceleration via Tempo-Enriched Policy and RL Fine-Tuning

ARC-RL: A Reinforcement Learning Playground Inspired by ARC Raiders

MRS: Multi-Resolution Skills for HRL Agents

Chain of Uncertain Rewards with Large Language Models for Reinforcement Learning

Multi-Camera View Scaling for Data-Efficient Robot Imitation Learning

ELMUR: External Layer Memory with Update/Rewrite for Long-Horizon RL Problems

Seeing Farther and Smarter: Value-Guided Multi-Path Reflection for VLM Policy Optimization

TWISTED-RL: Hierarchical Skilled Agents for Knot-Tying without Human Demonstrations

A Multi-Fidelity Control Variate Approach for Policy Gradient Estimation

VIKI-R: Coordinating Embodied Multi-Agent Cooperation via Reinforcement Learning

Autonomous Reinforcement Learning Robot Control with Intel's Loihi 2 Neuromorphic Hardware

Coarse-to-fine Q-Network with Action Sequence for Data-Efficient Reinforcement Learning

Using Non-Expert Data to Robustify Imitation Learning via Offline Reinforcement Learning

SimpleVLA-RL: Scaling VLA Training via Reinforcement Learning

PANAMA: A Network-Aware MARL Framework for Multi-Agent Path Finding in Digital Twin Ecosystems

ReCoDe: Reinforcement Learning-based Dynamic Constraint Design for Multi-Agent Coordination

Graph-Assisted Stitching for Offline Hierarchical Reinforcement Learning

DiffTORI: Differentiable Trajectory Optimization for Deep Reinforcement and Imitation Learning

VITAL: Interactive Few-Shot Imitation Learning via Visual Human-in-the-Loop Corrections

TWIST: Teleoperated Whole-Body Imitation System

Relational Object-Centric Actor-Critic

ES-Parkour: Advanced Robot Parkour with Bio-inspired Event Camera and Spiking Neural Network

Data Augmentation for Instruction Following Policies via Trajectory Segmentation

Towards Learning Scalable Agile Dynamic Motion Planning for Robosoccer Teams with Policy Optimization

Improving Vision-Language-Action Model with Online Reinforcement Learning

RAT: Adversarial Attacks on Deep Reinforcement Agents for Targeted Behaviors

Diffusion-Reward Adversarial Imitation Learning

Exploring the Edges of Latent State Clusters for Goal-Conditioned Reinforcement Learning

RIME: Robust Preference-based Reinforcement Learning with Noisy Preferences

FLaRe: Achieving Masterful and Adaptive Robot Policies with Large-Scale Reinforcement Learning Fine-Tuning