世界模型 - arXivDaily 专题

2407.20164 2024-07-30 cs.RO cs.AI cs.LG 60%

Language-Conditioned Offline RL for Multi-Robot Navigation

Steven Morad, Ajay Shankar, Jan Blumenkamp, Amanda Prorok

专题命中具身与机器人：environment model(abstract)；分类 cs.AI、cs.LG、cs.RO

2407.07889 2024-07-11 cs.RO cs.CV cs.LG 60%

AdaptiGraph: Material-Adaptive Graph-Based Neural Dynamics for Robotic Manipulation

Kaifeng Zhang, Baoyu Li, Kris Hauser, Yunzhu Li

专题命中具身与机器人：分类 cs.LG、cs.CV、cs.RO；dynamics model(abstract)；predictive model(abstract)；predictive models(abstract)

Comments Project page: https://robopil.github.io/adaptigraph/

2401.03306 2024-01-09 cs.LG cs.AI cs.RO 60%

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

专题命中具身与机器人：model-based RL(abstract)；分类 cs.AI、cs.LG、cs.RO

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

2207.12062 2023-10-24 cs.LG cs.RO 60%

Adaptive Asynchronous Control Using Meta-learned Neural Ordinary Differential Equations

Achkan Salehi, Steffen Rühl, Stephane Doncieux

专题命中具身与机器人：model-based reinforcement learning(abstract)；分类 cs.LG、cs.RO；dynamics model(abstract)

Comments 16 double column pages, 14 figures, 3 tables

Journal ref IEEE Transactions on Robotics; Print ISSN: 1552-3098; Online ISSN: 1941-0468

2309.08728 2023-09-19 cs.RO cs.AI 60%

SculptBot: Pre-Trained Models for 3D Deformable Object Manipulation

Alison Bartsch, Charlotte Avra, Amir Barati Farimani

专题命中具身与机器人：latent dynamics(abstract)；分类 cs.AI、cs.RO；dynamics model(abstract)

2305.02058 2023-05-04 cs.AI cs.HC cs.MA cs.RO 60%

Human Machine Co-adaption Interface via Cooperation Markov Decision Process System

Kairui Guo, Adrian Cheng, Yaqi Li, Jun Li, Rob Duffield, Steven W. Su

专题命中具身与机器人：model-based reinforcement learning(abstract)；分类 cs.AI、cs.RO、cs.MA

Comments 11 pages; 4 figures

2210.03113 2023-02-06 cs.RO cs.AI cs.LG 60%

IR-MCL: Implicit Representation-Based Online Global Localization

Haofei Kuang, Xieyuanli Chen, Tiziano Guadagnino, Nicky Zimmerman, Jens Behley, Cyrill Stachniss

专题命中具身与机器人：environment model(abstract)；分类 cs.AI、cs.LG、cs.RO

Comments 8 pages, 5 figures. Accepted to IEEE Robotics and Automation Letters

2210.10865 2022-10-21 cs.RO cs.AI cs.LG cs.SY eess.SY 60%

Robotic Table Wiping via Reinforcement Learning and Whole-body Trajectory Optimization

Thomas Lew, Sumeet Singh, Mario Prats, Jeffrey Bingham, Jonathan Weisz, Benjie Holson, Xiaohan Zhang, Vikas Sindhwani, Yao Lu, Fei Xia, Peng Xu, Tingnan Zhang, Jie Tan, Montserrat Gonzalez

专题命中具身与机器人：latent dynamics(abstract)；分类 cs.AI、cs.LG、cs.RO

2103.13842 2021-11-01 cs.RO cs.LG 60%

Model Predictive Actor-Critic: Accelerating Robot Skill Acquisition with Deep Reinforcement Learning

Andrew S. Morgan, Daljeet Nandha, Georgia Chalvatzaki, Carlo D'Eramo, Aaron M. Dollar, Jan Peters

专题命中具身与机器人：model-based reinforcement learning(abstract)；分类 cs.LG、cs.RO；dynamics model(abstract)

Comments IEEE International Conference on Robotics and Automation (ICRA), Xi'an, China, 2021

Journal ref 2021 IEEE International Conference on Robotics and Automation (ICRA)

2003.04663 2021-01-08 cs.RO cs.AI cs.LG 60%

Fast Online Adaptation in Robotics through Meta-Learning Embeddings of Simulated Priors

Rituraj Kaushik, Timothée Anne, Jean-Baptiste Mouret

专题命中具身与机器人：model-based reinforcement learning(abstract)；分类 cs.AI、cs.LG、cs.RO

Comments 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) | Video: http://tiny.cc/famle_video

2011.04999 2020-11-11 cs.RO cs.AI cs.LG 60%

Untangling Dense Knots by Learning Task-Relevant Keypoints

Jennifer Grannen, Priya Sundaresan, Brijen Thananjeyan, Jeffrey Ichnowski, Ashwin Balakrishna, Minho Hwang, Vainavi Viswanath, Michael Laskey, Joseph E. Gonzalez, Ken Goldberg

专题命中具身与机器人：environment model(abstract)；分类 cs.AI、cs.LG、cs.RO

Comments Conference on Robot Learning (CoRL) 2020 Oral. First two authors contributed equally

Journal ref 4th Conference on Robot Learning (CoRL 2020)

1912.04443 2020-06-23 cs.RO cs.CV cs.LG 60%

AVID: Learning Multi-Stage Tasks via Pixel-Level Translation of Human Videos

Laura Smith, Nikita Dhawan, Marvin Zhang, Pieter Abbeel, Sergey Levine

专题命中具身与机器人：model-based RL(abstract)；分类 cs.LG、cs.CV、cs.RO

Comments Robotics: Science and Systems (RSS) 2020 camera ready submission. Project website: https://sites.google.com/view/rss20avid

1907.03423 2020-05-19 cs.LG cs.AI cs.RO 60%

On-Policy Robot Imitation Learning from a Converging Supervisor

Ashwin Balakrishna, Brijen Thananjeyan, Jonathan Lee, Felix Li, Arsh Zahed, Joseph E. Gonzalez, Ken Goldberg

专题命中具身与机器人：model-based RL(abstract)；分类 cs.AI、cs.LG、cs.RO

Comments Conference on Robot Learning (CoRL) 2019 Oral. First two authors contributed equally

Journal ref 3rd Conference on Robot Learning (CoRL 2019)

1904.05538 2019-04-12 cs.RO cs.AI cs.LG 60%

Improvisation through Physical Understanding: Using Novel Objects as Tools with Visual Foresight

Annie Xie, Frederik Ebert, Sergey Levine, Chelsea Finn

专题命中具身与机器人：model-based RL(abstract)；分类 cs.AI、cs.LG、cs.RO

Comments Videos available at https://sites.google.com/view/gvf-tool

1903.09243 2019-03-25 cs.RO cs.AI cs.CL cs.LG 60%

Inferring Compact Representations for Efficient Natural Language Understanding of Robot Instructions

Siddharth Patki, Andrea F. Daniele, Matthew R. Walter, Thomas M. Howard

专题命中具身与机器人：environment model(abstract)；分类 cs.AI、cs.LG、cs.RO

Comments Accepted to ICRA 2019

1812.00568 2018-12-04 cs.RO cs.AI cs.CV cs.LG 60%

Visual Foresight: Model-Based Deep Reinforcement Learning for Vision-Based Robotic Control

Frederik Ebert, Chelsea Finn, Sudeep Dasari, Annie Xie, Alex Lee, Sergey Levine

专题命中具身与机器人：分类 cs.AI、cs.LG、cs.CV；predictive model(abstract)；predictive models(abstract)

1710.08893 2018-06-14 cs.RO cs.AI cs.LG 60%

Fast Model Identification via Physics Engines for Data-Efficient Policy Search

Shaojun Zhu, Andrew Kimmel, Kostas E. Bekris, Abdeslam Boularias

专题命中具身与机器人：model-based reinforcement learning(abstract)；分类 cs.AI、cs.LG、cs.RO

Comments IJCAI 18

1705.05172 2017-11-30 cs.LG cs.AI cs.HC cs.RO stat.ML 60%

Emotion in Reinforcement Learning Agents and Robots: A Survey

Thomas M. Moerland, Joost Broekens, Catholijn M. Jonker

专题命中具身与机器人：model-based RL(abstract)；分类 cs.AI、cs.LG、cs.RO

Comments To be published in Machine Learning Journal

Journal ref Machine Learning 2017

1509.06841 2016-08-12 cs.LG cs.RO 60%

One-Shot Learning of Manipulation Skills with Online Dynamics Adaptation and Neural Network Priors

Justin Fu, Sergey Levine, Pieter Abbeel

专题命中具身与机器人：model-based reinforcement learning(abstract)；分类 cs.LG、cs.RO；dynamics model(abstract)

2503.11999 2025-09-03 cs.RO cs.CV cs.SY eess.SY 58%

Diffusion Dynamics Models with Generative State Estimation for Cloth Manipulation

Tongxuan Tian, Haoyang Li, Bo Ai, Xiaodi Yuan, Zhiao Huang, Hao Su

机构 * University of California San Diego（加州大学圣地亚哥分校）； Hillbot Inc（Hillbot公司）

专题命中具身与机器人：dynamics model(title,abstract)；分类 cs.CV、cs.RO

Comments CoRL 2025. Project website: https://uniclothdiff.github.io/

2503.23271 2025-04-01 cs.RO cs.AI 58%

Learning Coordinated Bimanual Manipulation Policies using State Diffusion and Inverse Dynamics Models

Haonan Chen, Jiaming Xu, Lily Sheng, Tianchen Ji, Shuijing Liu, Yunzhu Li, Katherine Driggs-Campbell

专题命中具身与机器人：dynamics model(title,abstract)；分类 cs.AI、cs.RO

Comments Project Page: https://haonan16.github.io/coord_bimanual_page/. 12 pages, 12 figures, Accepted at ICRA 2025

2307.09206 2024-10-08 cs.RO cs.LG 58%

Context-Conditional Navigation with a Learning-Based Terrain- and Robot-Aware Dynamics Model

Suresh Guttikonda, Jan Achterhold, Haolong Li, Joschka Boedecker, Joerg Stueckler

专题命中具身与机器人：dynamics model(title,abstract)；分类 cs.LG、cs.RO

2409.11452 2024-09-19 cs.RO cs.LG 58%

Learning a Terrain- and Robot-Aware Dynamics Model for Autonomous Mobile Robot Navigation

Jan Achterhold, Suresh Guttikonda, Jens U. Kreber, Haolong Li, Joerg Stueckler

专题命中具身与机器人：dynamics model(title,abstract)；分类 cs.LG、cs.RO

Comments Submitted to Robotics and Autonomous Systems. arXiv admin note: substantial text overlap with arXiv:2307.09206

2311.00802 2023-11-03 cs.RO cs.LG 58%

Neural Field Dynamics Model for Granular Object Piles Manipulation

Shangjie Xue, Shuo Cheng, Pujith Kachana, Danfei Xu

专题命中具身与机器人：dynamics model(title,abstract)；分类 cs.LG、cs.RO

2202.12977 2022-07-19 cs.RO cs.AI 58%

Learning physics-informed simulation models for soft robotic manipulation: A case study with dielectric elastomer actuators

Manu Lahariya, Craig Innes, Chris Develder, Subramanian Ramamoorthy

专题命中具身与机器人：simulation model(title)；分类 cs.AI、cs.RO；dynamics model(abstract)

2204.08647 2022-04-22 cs.RO cs.LG 58%

Learning Forward Dynamics Model and Informed Trajectory Sampler for Safe Quadruped Navigation

Yunho Kim, Chanyoung Kim, Jemin Hwangbo

专题命中具身与机器人：dynamics model(title,abstract)；分类 cs.LG、cs.RO

Comments Accepted to Robotics: Science and Systems (RSS 2022)

1909.11652 2019-09-26 cs.RO cs.LG 58%

Deep Dynamics Models for Learning Dexterous Manipulation

Anusha Nagabandi, Kurt Konoglie, Sergey Levine, Vikash Kumar

专题命中具身与机器人：dynamics model(title,abstract)；分类 cs.LG、cs.RO

Comments project website https://sites.google.com/view/pddm/

2606.31941 2026-07-01 cs.RO cs.AI 新提交 56%

LeCropFollow: Latent Space Planning for Navigation in Unstructured Crop Fields

LeCropFollow：非结构化农田中导航的潜在空间规划

Felipe Tommaselli, Francisco Affonso, Arthur Pompeu, Gianluca Capezzuto, Arun Narenthiran Sivakumar, Girish Chowdhary, Marcelo Becker

机构 * University of São Paulo（圣保罗大学）； University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）

专题命中具身与机器人：model-based reinforcement learning(abstract)；分类 cs.AI、cs.RO

AI总结提出LeCropFollow框架，利用自监督语义热图和基于模型的强化学习规划器在潜在流形中优化轨迹，无需几何建模，实现从简化模拟到真实世界的零样本迁移，在玉米田间隙中语义故障减少2.4倍。

Comments 8 pages, 7 figures, 3 tables. Github Repo: https://github.com/Felipe-Tommaselli/lecropfollow

Journal ref IEEE Robotics and Automation Letters, 2026

1903.02531 2026-06-04 cs.RO cs.AI cs.CV cs.LG cs.SY eess.SY 56%

Combining Optimal Control and Learning for Visual Navigation in Novel Environments

将最优控制与学习相结合用于新环境中的视觉导航

Somil Bansal, Varun Tolani, Saurabh Gupta, Jitendra Malik, Claire Tomlin

机构 * University of California, Berkeley（加州大学伯克利分校）； Facebook AI Research（脸书人工智能研究）

专题命中具身与机器人：分类 cs.AI、cs.LG、cs.CV；dynamics model(abstract)

AI总结本文提出了一种结合模型控制与学习感知的方法，用于在新环境中实现可靠的视觉导航，通过生成无碰撞路径的 waypoints，使机器人能够高效地到达目标位置，同时在低帧率和仿真到现实的迁移中表现良好。

Comments Project website: https://vtolani95.github.io/WayPtNav/

1808.00113 2026-06-04 eess.SY cs.LG cs.RO cs.SY math.OC 56%

Learning Stabilizable Dynamical Systems via Control Contraction Metrics

通过控制收缩度量学习可稳定化的动态系统

Sumeet Singh, Vikas Sindhwani, Jean-Jacques E. Slotine, Marco Pavone

机构 * Dept. of Aeronautics and Astronautics, Stanford University（航空航天系，斯坦福大学）； Google Brain Robotics, New York（谷歌大脑机器人，纽约）； Dept. of Mechanical Engineering, Massachusetts Institute of Technology（机械工程系，麻省理工学院）

专题命中具身与机器人：model-based reinforcement learning(abstract)；分类 cs.LG、cs.RO

AI总结本文提出了一种新的框架，用于学习可稳定化的非线性动态系统，以实现机器人连续控制任务。核心方法是开发一种基于稳定性的控制理论正则化器，以确保学习到的系统可以配备一个稳健的控制器，能够稳定任何系统生成的开环轨迹。通过利用收缩理论、统计学习和凸优化工具，我们提供了一个通用且可操作的半监督算法来学习可稳定化的动态系统，可以应用于复杂的欠驱动系统。在模拟平面四旋翼系统上验证了所提算法，并观察到与传统回归技术学习的模型相比，使用控制理论正则化模型在轨迹生成和跟踪性能上有显著改进，尤其是在使用少量示范示例时。结果展示了将标准基于模型的强化学习算法与非线性控制理论概念结合的必要性，以提高可靠性。

Comments To appear at WAFR 2018. v2: re-structured Sections 3 & 4 to improve clarity; expanded discussion on limitations & future work in Section 5; added details on training & validation, significantly expanded experiments

1. 具身与机器人 510 篇

Language-Conditioned Offline RL for Multi-Robot Navigation

AdaptiGraph: Material-Adaptive Graph-Based Neural Dynamics for Robotic Manipulation

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

Adaptive Asynchronous Control Using Meta-learned Neural Ordinary Differential Equations

SculptBot: Pre-Trained Models for 3D Deformable Object Manipulation

Human Machine Co-adaption Interface via Cooperation Markov Decision Process System

IR-MCL: Implicit Representation-Based Online Global Localization

Robotic Table Wiping via Reinforcement Learning and Whole-body Trajectory Optimization

Model Predictive Actor-Critic: Accelerating Robot Skill Acquisition with Deep Reinforcement Learning

Fast Online Adaptation in Robotics through Meta-Learning Embeddings of Simulated Priors

Untangling Dense Knots by Learning Task-Relevant Keypoints

AVID: Learning Multi-Stage Tasks via Pixel-Level Translation of Human Videos

On-Policy Robot Imitation Learning from a Converging Supervisor

Improvisation through Physical Understanding: Using Novel Objects as Tools with Visual Foresight

Inferring Compact Representations for Efficient Natural Language Understanding of Robot Instructions

Visual Foresight: Model-Based Deep Reinforcement Learning for Vision-Based Robotic Control

Fast Model Identification via Physics Engines for Data-Efficient Policy Search

Emotion in Reinforcement Learning Agents and Robots: A Survey

One-Shot Learning of Manipulation Skills with Online Dynamics Adaptation and Neural Network Priors

Diffusion Dynamics Models with Generative State Estimation for Cloth Manipulation

Learning Coordinated Bimanual Manipulation Policies using State Diffusion and Inverse Dynamics Models

Context-Conditional Navigation with a Learning-Based Terrain- and Robot-Aware Dynamics Model

Learning a Terrain- and Robot-Aware Dynamics Model for Autonomous Mobile Robot Navigation

Neural Field Dynamics Model for Granular Object Piles Manipulation

Learning physics-informed simulation models for soft robotic manipulation: A case study with dielectric elastomer actuators

Learning Forward Dynamics Model and Informed Trajectory Sampler for Safe Quadruped Navigation

Deep Dynamics Models for Learning Dexterous Manipulation

LeCropFollow: Latent Space Planning for Navigation in Unstructured Crop Fields

Combining Optimal Control and Learning for Visual Navigation in Novel Environments

Learning Stabilizable Dynamical Systems via Control Contraction Metrics