cs.RO（2026-10-20）

📊 共 42 篇论文 | 🔗 7 篇有代码

🎯 兴趣领域导航

支柱一：机器人控制 (Robot Control) (31 🔗6) 支柱九：具身大模型 (Embodied Foundation Models) (5) 支柱二：RL算法与架构 (RL & Architecture) (3 🔗1) 支柱八：物理动画 (Physics-based Animation) (1) 支柱三：空间感知与语义 (Perception & Semantics) (1) 支柱六：视频提取与匹配 (Video Extraction) (1)

🔬 支柱一：机器人控制 (Robot Control) (31 篇)

#题目一句话要点标签🔗⭐
1 Fast Motion Retargeting Pipeline for Scalable Humanoid Robot Learning 提出FRDV框架以解决人形机器人运动重定向问题 humanoid humanoid robot motion tracking ✅
2 SuperMimic: Scaling Humanoid Whole-Body Control with One Differentiable World Model 提出SuperMimic以解决人形机器人全身控制的计算瓶颈问题 humanoid whole-body control motion tracking
3 LocoPilot: Single-Operator Bilateral Teleoperation for Loco-Manipulation Policy Learning 提出LocoPilot以解决单操作者双向遥控的挑战 quadruped manipulation loco-manipulation ✅
4 Adaptive Traversal Tubes for Posture-Aware Navigation of Bipedal Humanoids 提出自适应通行管道以解决双足人形机器人姿态感知导航问题 humanoid bipedal biped
5 RoboCasa-Shift: Stage-Aware Robustness Evaluation of Vision-Language-Action Policies in Long-Horizon Household Manipulation 提出RoboCasa-Shift以解决长时间家居操作中的VLA鲁棒性评估问题 manipulation vision-language-action VLA ✅
6 VLM-Guided Contact Retargeting and Generative Partner Modeling for Humanoid-Human Interaction 提出REACT框架以解决人形机器人与人类交互中的部分可观测性问题 humanoid Unitree egocentric
7 Physical Canvas: A Vision Tactile World Model For Robot Policy Evaluation And Learning 提出PhysCanvas以解决机器人政策评估中的接触动态问题 manipulation world model world models ✅
8 Reasoning as a Distribution: Generative Plan Latents for Vision-Language-Action Models 提出生成计划潜变量以解决视觉-语言-行动模型的推理问题 manipulation vision-language-action VLA
9 From Predictive Geometry to Action: Learning Complementary Predictive Representations for Embodied Manipulation 提出双阶段框架以解决动作生成中的预测几何表示问题 manipulation bi-manual VLA ✅
10 FIT-MAC: A Closed-Loop Multi-Agent Framework for Robust Vision-Language-Action Models 提出FIT-MAC框架以提升视觉-语言-动作模型的鲁棒性 manipulation vision-language-action VLA
11 Learning Adaptable Dexterous Skills via Hybrid Dexterous Mimicking 提出AdaDex框架以解决灵巧技能学习中的适应性问题 motion tracking reinforcement learning HOI
12 Frontier LLMs Enter the World of Robotics: A Cross-Model Evaluation on RoboColiseum 通过Agent2Act评估多模态大语言模型在机器人控制中的应用 manipulation dexterous hand large language model
13 PLaCA: Learning Patch-Level Future Latent Change for Action Generation 提出PLaCA以解决机器人动作生成中的环境变化预测问题 manipulation latent dynamics vision-language-action
14 AWLS: Advantage-Weighted Latent Steering for One-Step Generative Robotic Manipulation Policies 提出AWLS以优化一阶段生成机器人操作策略 manipulation offline RL behavior cloning
15 Learning Tactile-Aware Hand Representations from Human Hand-Object Interaction Data 提出HuTact框架以利用人手物体交互数据提升机器人触觉学习 manipulation dexterous manipulation policy learning
16 Large Multimodal Models for Direct Robot Control: GPT-6 as a Beginning 提出GPT-6 Astra以解决机器人直接控制中的能力不足问题 manipulation multimodal
17 Plausible Future, Wrong Choice: Steering World-Action Models toward Intended Future 提出TaskWAM以解决长时间操作中的阶段混淆问题 humanoid humanoid robot manipulation
18 Real-Time Disturbance-Aware Koopman Model Predictive Control for Quadrotors in Spatially Varying Winds 提出一种实时干扰感知的Koopman模型预测控制以解决四旋翼在变化风场中的飞行问题 MPC model predictive control predictive model
19 EventSync: Event-Driven Memory Updates for History-Dependent Manipulation 提出EventSync以解决历史依赖机器人操控中的记忆更新问题 manipulation vision-language-action
20 Prescribed Performance Guided Safe Reinforcement Learning for Roundabout Navigation of Autonomous Vehicles 提出RLTrajGen-PPMPC-CBF框架以解决自动驾驶车辆环形交叉口导航问题 MPC model predictive control reinforcement learning
21 SceneRig: An Agentic System for Simulation-Ready 3D Scene Reconstruction from Single Images 提出SceneRig以解决单图像3D场景重建问题 manipulation scene reconstruction
22 Physical AI for Intralogistics: Benchmarking the NVIDIA Sim-to-Real Ecosystem for Cosmos-Enhanced COMPASS Navigation on Industrial Mobile Robots 基于NVIDIA Cosmos的COMPASS导航提升工业移动机器人导航性能 humanoid sim-to-real
23 EchoNav: Evidential BEV Completion from Historical Observations for Visual Navigation 提出EchoNav框架以解决视觉导航中的障碍物推断问题 quadruped humanoid humanoid robot
24 UniWALM: Unified World Action Language Model for Embodied Robot Control 提出UniWALM以解决机器人控制中的多模态学习问题 manipulation flow matching VLA
25 AgenticRoboRL: Bringing Coding-Agent Post-Training to Agentic Robotics 提出AgenticRoboRL以提升机器人规划能力 manipulation reinforcement learning VLA
26 Embodied Task Agents Can Perform Fine Manipulation through Tactile In-Context Learning 提出Tactile-ETA以解决细致操作中的触觉学习问题 manipulation
27 HATCH: Hardware-Attested Task Chains for Verifiable Manipulation 提出HATCH以解决机器人操作能力验证问题 manipulation
28 GPT-6-Astra Understands Touch and Outperform Vision--Tactile--Language--Action Models 提出GPT-6-Astra以解决触觉理解与机器人操控问题 manipulation multimodal
29 What Novice Operators See Changes What the Policy Learns: A User Study of Keypoint Hand Visual Representation for Teleoperated Demonstration 提出基于关键点手部视觉表示的用户研究以提升模仿学习效果 teleoperation imitation learning diffusion policy ✅
30 Towards Autonomous Bricklaying Using Boston Dynamics Spot 提出基于Boston Dynamics Spot的自主砌砖系统以解决建筑施工问题 quadruped legged robot
31 Temporal Difference Learning of Probabilistic Reach-Avoid Value Functions 提出概率性达避值函数以解决不确定环境中的任务安全问题 locomotion

🔬 支柱九：具身大模型 (Embodied Foundation Models) (5 篇)

#题目一句话要点标签🔗⭐
32 PACT-VLA: Speculative Action-Chunk Verification with Pre-Commit Observations 提出PACT-VLA以解决视觉-语言-动作策略中的推测性动作块验证问题 vision-language-action VLA
33 CovNav: Conformal Prediction-Guided Adaptive Inference for Object Goal Navigation 提出CovNav以解决物体目标导航中的推理效率问题 large language model chain-of-thought
34 CAPSafe: A Benchmark for Contextual Action and Plan Safety Reasoning in Embodied Agents 提出CAPSafe基准以解决嵌入式智能体的安全推理问题 large language model
35 JARVIS: From Visual Manuals to Autonomous Robotic Assembly 提出JARVIS以解决机器人自主组装中的视觉手册执行问题 multimodal
36 Evaluating Social Reasoning in Embodied Vision-Language Models 评估具身视觉语言模型的社会推理能力 foundation model

🔬 支柱二：RL算法与架构 (RL & Architecture) (3 篇)

#题目一句话要点标签🔗⭐
37 Perceives but Doesn't Brake: A Perception-Action Binding Gap in Reasoning Driving VLAs 提出一种方法以解决驾驶模型中的感知-行动绑定缺口问题 flow matching vision-language-action VLA
38 Multi-View Latent Action Pretraining for Cross-Embodiment Robot Learning from Videos 提出多视角潜在动作预训练以解决跨体态机器人学习问题 flow matching cross-embodiment ✅
39 Knowing When to Stop Thinking: A Dissection of Expected Free Energy as a Test-Time Compute Controller for Latent World-Model Planning 提出基于期望自由能量的计算控制方法以优化潜在世界模型规划 world model world models

🔬 支柱八：物理动画 (Physics-based Animation) (1 篇)

#题目一句话要点标签🔗⭐
40 Industrial ReMEmbR: Scaling Spatiotemporal Memory to Robot Fleets 提出Industrial ReMEmbR以解决机器人车队的时空记忆问题 spatiotemporal

🔬 支柱三：空间感知与语义 (Perception & Semantics) (1 篇)

#题目一句话要点标签🔗⭐
41 MARE: An Open-Source Lunar Rover Simulator with Shadow Ground Truth for Perception and Autonomy 提出MARE以解决月球探测中的感知与自主导航问题 stereo depth elevation map

🔬 支柱六：视频提取与匹配 (Video Extraction) (1 篇)

#题目一句话要点标签🔗⭐
42 A Head Camera Is (Almost) All You Need: An Electronics-Free UMI Glove Labelled and Observed by One Moving Camera 提出一种基于头部摄像头的无电子手套标记与观察方法 egocentric

⬅️ 返回 cs.RO 首页 · 🏠 返回主页