机器人 / 具身智能 - arXivDaily 专题

2509.19080 2026-03-23 cs.RO cs.AI 92%

World4RL: Diffusion World Models for Policy Refinement with Reinforcement Learning for Robotic Manipulation

World4RL: 基于扩散世界模型的强化学习政策精修框架用于机器人操作

Zhennan Jiang, Kai Liu, Yuxin Qin, Shuai Tian, Yupeng Zheng, Mingcai Zhou, Chao Yu, Haoran Li, Dongbin Zhao

机构 * The State Key Laboratory of Multimodal Artificial Intelligence Systems, Institute of Automation, Chinese Academy of Sciences（多模态人工智能系统国家重点实验室，自动化研究所，中国科学院）； School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院）； Zhongguancun Academy（中关村学院）； Beijing Zhongke Huiling Robot Technology Co（北京中科创联机器人技术有限公司）； Department of Electronic Engineering, Tsinghua University（清华大学电子工程系）

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

AI总结 World4RL通过扩散世界模型提升机器人操作政策的精修效果，采用高保真模拟环境进行端到端政策优化，优于模仿学习及其他基线方法。

2501.10100 2025-12-16 cs.RO cs.AI cs.LG 91%

Robotic World Model: A Neural Network Simulator for Robust Policy Optimization in Robotics

机器人世界模型：一种用于机器人鲁棒策略优化的神经网络模拟器

Chenhao Li, Andreas Krause, Marco Hutter

机构 * ETH Zurich（苏黎世联邦理工学院）

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

AI总结本文提出了一种基于神经网络的机器人世界模型，通过双自回归机制和自监督训练实现鲁棒的长周期预测，从而提升机器人策略优化的效率和适应性。

2510.14830 2026-03-11 cs.RO cs.AI cs.LG 90%

RL-100: Performant Robotic Manipulation with Real-World Reinforcement Learning

RL-100：基于现实强化学习的高效机器人操作

Kun Lei, Huanyu Li, Dongjie Yu, Zhenyu Wei, Lingxiao Guo, Zhennan Jiang, Ziyu Wang, Shiyu Liang, Huazhe Xu

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

AI总结 RL-100通过结合扩散视觉-运动策略和截断PPO目标，实现了高效且鲁棒的现实机器人操作，展示了在多种任务中高达100%的成功率和良好的适应性。

Comments https://lei-kun.github.io/RL-100/

2504.16680 2026-01-09 cs.RO cs.AI cs.LG 90%

Uncertainty-Aware Robotic World Model Makes Offline Model-Based Reinforcement Learning Work on Real Robots

具有不确定性的机器人世界模型使基于模型的离线强化学习在真实机器人上有效

Chenhao Li, Andreas Krause, Marco Hutter

机构 * ETH Zurich（苏黎世联邦理工学院）

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

AI总结本文提出了一种具有不确定性的机器人世界模型，结合MOPO-PPO框架，使基于模型的离线强化学习在真实机器人上有效，通过提高策略的鲁棒性并超越现有基线。

2503.01837 2025-10-06 cs.LG cs.CV cs.RO 90%

Multi-Stage Manipulation with Demonstration-Augmented Reward, Policy, and World Model Learning

Adrià López Escoriza, Nicklas Hansen, Stone Tao, Tongzhou Mu, Hao Su

机构 * University of California San Diego（加州大学圣迭戈分校）； ETH Zürich, Switzerland（苏黎世联邦理工学院）

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

Comments Project page can be found at https://adrialopezescoriza.github.io/demo3/

Journal ref ICML 2025

2507.13277 2025-07-18 cs.RO cs.AI cs.LG 90%

Evaluating Reinforcement Learning Algorithms for Navigation in Simulated Robotic Quadrupeds: A Comparative Study Inspired by Guide Dog Behaviour

Emma M. A. Harrison

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

2505.18719 2025-05-27 cs.RO cs.AI 90%

VLA-RL: Towards Masterful and General Robotic Manipulation with Scalable Reinforcement Learning

Guanxing Lu, Wenkai Guo, Chubin Zhang, Yuheng Zhou, Haonan Jiang, Zifeng Gao, Yansong Tang, Ziwei Wang

机构 * Tsinghua Shenzhen International Graduate School, Tsinghua University（清华大学深圳国际研究生院，清华大学）； School of Electrical and Electronic Engineering, Nanyang Technological University（南洋理工大学电子与电气工程学院）

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

2205.03353 2022-05-09 cs.RO cs.LG 90%

How to Spend Your Robot Time: Bridging Kickstarting and Offline Reinforcement Learning for Vision-based Robotic Manipulation

Alex X. Lee, Coline Devin, Jost Tobias Springenberg, Yuxiang Zhou, Thomas Lampe, Abbas Abdolmaleki, Konstantinos Bousmalis

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

2010.13766 2022-02-14 cs.RO cs.LG 90%

Contextual Latent-Movements Off-Policy Optimization for Robotic Manipulation Skills

Samuele Tosatto, Georgia Chalvatzaki, Jan Peters

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

2110.10905 2021-10-22 cs.RO cs.AI 90%

Efficient Robotic Manipulation Through Offline-to-Online Reinforcement Learning and Goal-Aware State Information

Jin Li, Xianyuan Zhan, Zixu Xiao, Guyue Zhou

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

2607.13033 2026-07-15 cs.RO 新提交 89%

DenseReward: Dense Reward Learning via Failure Synthesis for Robotic Manipulation

DenseReward：通过失败合成进行密集奖励学习以实现机器人操作

Yu Fang, Wanxi Dong, Jiaqi Liu, Yue Yang, Mingxiao Huo, Yao Mu, Huaxiu Yao, Li Erran Li, Daniel Szafir, Mingyu Ding

机构 * University of North Carolina at Chapel Hill（北卡罗来纳大学教堂山分校）； Carnegie Mellon University（卡内基梅隆大学）； Shanghai Jiao Tong University（上海交通大学）； Amazon AWS AI（亚马逊AWS人工智能）

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

AI总结研究针对强化学习中缺乏可靠奖励模型的问题，提出DenseReward模型，通过自动生成失败数据合成逼真轨迹，从视觉和语言预测密集奖励分数，在模拟和现实操作中表现优异，还为下游任务提供指导并发布相关资源。

Comments Website: https://dense-reward.github.io/

2504.21769 2025-05-01 cs.RO 89%

LLM-based Interactive Imitation Learning for Robotic Manipulation

Jonas Werner, Kun Chu, Cornelius Weber, Stefan Wermter

机构 * Knowledge Technology, Department of Informatics University of Hamburg（信息学院知识技术部汉堡大学）

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

Comments To be published in IJCNN 2025 proceedings

2210.08126 2023-09-15 cs.RO 89%

Geometric Reinforcement Learning For Robotic Manipulation

Naseem Alhousani, Matteo Saveriano, Ibrahim Sevinc, Talha Abdulkuddus, Hatice Kose, Fares J. Abu-Dakka

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

Comments 14 pages, 14 figures, journal

2603.27346 2026-04-03 cs.RO cs.AI cs.LG 89%

D-SPEAR: Dual-Stream Prioritized Experience Adaptive Replay for Stable Reinforcement Learning in Robotic Manipulation

D-SPEAR：双流优先经验自适应回放用于稳定机器人操作的强化学习

Yu Zhang, Karl Mason

机构 * School of Computer Science University of Galway Galway, Ireland

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

AI总结本文提出D-SPEAR双流优先经验自适应回放框架，通过分离actor和critic采样并维护共享回放缓冲区，利用优先回放提升价值学习效率，低误差过渡稳定策略优化，在机器人操作任务中优于SAC、TD3等基线方法。

Comments Accepted at IEEE 11th International Conference on Control and Robotics Engineering (ICCRE 2026)

2107.13356 2021-07-29 cs.RO cs.AI 89%

Value-Based Reinforcement Learning for Continuous Control Robotic Manipulation in Multi-Task Sparse Reward Settings

Sreehari Rammohan, Shangqun Yu, Bowen He, Eric Hsiung, Eric Rosen, Stefanie Tellex, George Konidaris

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

Comments 5 pages, 2 figures, published at RSS 2021 workshop: Advancing Artificial Intelligence and Manipulation for Robotics: Understanding Gaps, Industry and Academic Perspectives, and Community Building

2503.04280 2025-06-11 cs.RO cs.AI cs.LG 89%

Towards Autonomous Reinforcement Learning for Real-World Robotic Manipulation with Large Language Models

Niccolò Turcato, Matteo Iovino, Aris Synodinos, Alberto Dalla Libera, Ruggero Carli, Pietro Falco

机构 * Department of Information Engineering, University of Padova（帕多瓦大学信息工程系）； ABB Corporate Research（ABB企业研究）

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

2505.08376 2025-05-14 cs.RO cs.AI cs.LG 89%

Adaptive Diffusion Policy Optimization for Robotic Manipulation

Huiyun Jiang, Zhuang Yang

机构 * School of Computer Science and Technology, Soochow University（计算机科学与技术学院，苏州大学）

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

2307.07091 2024-07-16 cs.LG cs.AI cs.RO 89%

Robotic Manipulation Datasets for Offline Compositional Reinforcement Learning

Marcel Hussing, Jorge A. Mendez, Anisha Singrodia, Cassandra Kent, Eric Eaton

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

Comments Published as a conference paper at the First Reinforcement Learning Conference (RLC)

2405.04549 2024-05-09 cs.CV cs.AI cs.RO 89%

ClothPPO: A Proximal Policy Optimization Enhancing Framework for Robotic Cloth Manipulation with Observation-Aligned Action Spaces

Libing Yang, Yang Li, Long Chen

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

1910.07294 2021-11-12 cs.LG cs.AI cs.RO stat.ML 89%

Reinforcement Learning for Robotic Manipulation using Simulated Locomotion Demonstrations

Ozsel Kilinc, Giovanni Montana

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

Comments To appear in ECML PKDD 2022

2604.28192 2026-05-08 cs.RO cs.CV 88%

LaST-R1: Reinforcing Robotic Manipulation via Adaptive Physical Latent Reasoning

LaST-R1：通过自适应物理潜在推理强化机器人操作

Hao Chen, Jiaming Liu, Zhonghao Yan, Nuowei Han, Renrui Zhang, Chenyang Gu, Jialin Gao, Ziyu Guo, Siyuan Qian, Yinxi Wang, Peng Jia, Shanghang Zhang, Pheng-Ann Heng

机构 * The Chinese University of Hong Kong（香港中文大学）； State Key Laboratory of Multimedia Information Processing, School of Computer Science, Peking University（北京大学多媒体信息处理国家重点实验室，计算机学院）； Simplexity Robotics Project（Simplexity机器人项目）

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

AI总结本文提出LaST-R1，一种基于强化学习的后训练框架，通过联合优化潜在推理过程和动作生成，提升机器人在动态环境中的适应性和泛化能力。

2509.11125 2026-03-13 cs.RO cs.CV 88%

ManiVID-3D: Generalizable View-Invariant Reinforcement Learning for Robotic Manipulation via Disentangled 3D Representations

ManiVID-3D: 通用的视角不变强化学习用于机器人操作的解耦3D表示

Zheng Li, Pei Qu, Yufei Jia, Shihui Zhou, Haizhou Ge, Jiahang Cao, Jinni Zhou, Guyue Zhou, Jun Ma

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

AI总结 ManiVID-3D通过解耦3D表示实现视角不变的强化学习，提升机器人操作在视角变化下的鲁棒性和效率。

Comments Accepted to RA-L. Project website: https://zheng-joe-lee.github.io/manivid3d/

2508.11143 2026-03-02 cs.RO cs.AI 88%

Actor-Critic for Continuous Action Chunks: A Reinforcement Learning Framework for Long-Horizon Robotic Manipulation with Sparse Reward

用于连续动作片段的Actor-Critic：一种针对长周期机械臂操作的强化学习框架

Jiarui Yang, Bin Zhu, Jingjing Chen, Yu-Gang Jiang

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

AI总结 AC3是一种用于连续动作分块的强化学习框架，通过定向稳定机制和自监督模块，在稀疏奖励环境下实现高效稳定的动作序列学习。

Comments 14 pages, 13 figures, Accepted by AAAI 2026 (oral)

2510.26406 2025-10-31 cs.RO cs.AI 88%

Human-in-the-loop Online Rejection Sampling for Robotic Manipulation

Guanxing Lu, Rui Zhao, Haitao Lin, He Zhang, Yansong Tang

机构 * Tsinghua Shenzhen International Graduate School（清华大学深圳国际研究生院）； Tencent Robotics X（腾讯机器人实验室）

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

Comments 8 pages

2304.06055 2025-09-23 cs.RO cs.AI 88%

Sample-Efficient Reinforcement Learning with Symmetry-Guided Demonstrations for Robotic Manipulation

Amir M. Soufi Enayati, Zengjie Zhang, Kashish Gupta, Homayoun Najjaran

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

2509.00319 2025-09-03 cs.RO cs.AI cs.SY eess.SY 88%

Contact-Aided Navigation of Flexible Robotic Endoscope Using Deep Reinforcement Learning in Dynamic Stomach

Chi Kit Ng, Huxin Gao, Tian-Ao Ren, Jiewen Lai, Hongliang Ren

机构 * Department of Electronic Engineering, The Chinese University of Hong Kong, Shatin, N.T., Hong Kong, China（电子工程系，香港中文大学（深圳）沙田分校，香港，中国）； Department of Mechanical Engineering, Stanford University（机械工程系，斯坦福大学）

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

2409.15688 2025-04-01 cs.RO cs.AI 88%

Safe Navigation for Robotic Digestive Endoscopy via Human Intervention-based Reinforcement Learning

Min Tan, Yushun Tao, Boyun Zheng, GaoSheng Xie, Lijuan Feng, Zeyang Xia, Jing Xiong

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

2412.10096 2024-12-16 cs.RO cs.LG 88%

Reward Machine Inference for Robotic Manipulation

Mattijs Baert, Sam Leroux, Pieter Simoens

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

2312.07953 2023-12-15 cs.RO cs.LG 88%

Enhancing Robotic Navigation: An Evaluation of Single and Multi-Objective Reinforcement Learning Strategies

Vicki Young, Jumman Hossain, Nirmalya Roy

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

Comments REU program project (work in progress)

2311.11287 2023-11-21 cs.RO cs.AI 88%

Tactile Active Inference Reinforcement Learning for Efficient Robotic Manipulation Skill Acquisition

Zihao Liu, Xing Liu, Yizhai Zhang, Zhengxiong Liu, Panfeng Huang

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

1. 模仿学习与强化学习 4103 篇

World4RL: Diffusion World Models for Policy Refinement with Reinforcement Learning for Robotic Manipulation

Robotic World Model: A Neural Network Simulator for Robust Policy Optimization in Robotics

RL-100: Performant Robotic Manipulation with Real-World Reinforcement Learning

Uncertainty-Aware Robotic World Model Makes Offline Model-Based Reinforcement Learning Work on Real Robots

Multi-Stage Manipulation with Demonstration-Augmented Reward, Policy, and World Model Learning

Evaluating Reinforcement Learning Algorithms for Navigation in Simulated Robotic Quadrupeds: A Comparative Study Inspired by Guide Dog Behaviour

VLA-RL: Towards Masterful and General Robotic Manipulation with Scalable Reinforcement Learning

How to Spend Your Robot Time: Bridging Kickstarting and Offline Reinforcement Learning for Vision-based Robotic Manipulation

Contextual Latent-Movements Off-Policy Optimization for Robotic Manipulation Skills

Efficient Robotic Manipulation Through Offline-to-Online Reinforcement Learning and Goal-Aware State Information

DenseReward: Dense Reward Learning via Failure Synthesis for Robotic Manipulation

LLM-based Interactive Imitation Learning for Robotic Manipulation

Geometric Reinforcement Learning For Robotic Manipulation

D-SPEAR: Dual-Stream Prioritized Experience Adaptive Replay for Stable Reinforcement Learning in Robotic Manipulation

Value-Based Reinforcement Learning for Continuous Control Robotic Manipulation in Multi-Task Sparse Reward Settings

Towards Autonomous Reinforcement Learning for Real-World Robotic Manipulation with Large Language Models

Adaptive Diffusion Policy Optimization for Robotic Manipulation

Robotic Manipulation Datasets for Offline Compositional Reinforcement Learning

ClothPPO: A Proximal Policy Optimization Enhancing Framework for Robotic Cloth Manipulation with Observation-Aligned Action Spaces

Reinforcement Learning for Robotic Manipulation using Simulated Locomotion Demonstrations

LaST-R1: Reinforcing Robotic Manipulation via Adaptive Physical Latent Reasoning

ManiVID-3D: Generalizable View-Invariant Reinforcement Learning for Robotic Manipulation via Disentangled 3D Representations

Actor-Critic for Continuous Action Chunks: A Reinforcement Learning Framework for Long-Horizon Robotic Manipulation with Sparse Reward

Human-in-the-loop Online Rejection Sampling for Robotic Manipulation

Sample-Efficient Reinforcement Learning with Symmetry-Guided Demonstrations for Robotic Manipulation

Contact-Aided Navigation of Flexible Robotic Endoscope Using Deep Reinforcement Learning in Dynamic Stomach

Safe Navigation for Robotic Digestive Endoscopy via Human Intervention-based Reinforcement Learning

Reward Machine Inference for Robotic Manipulation

Enhancing Robotic Navigation: An Evaluation of Single and Multi-Objective Reinforcement Learning Strategies

Tactile Active Inference Reinforcement Learning for Efficient Robotic Manipulation Skill Acquisition