cs.RO（2026-10-27）

📊 共 48 篇论文 | 🔗 7 篇有代码

🎯 兴趣领域导航

支柱一：机器人控制 (Robot Control) (35 🔗4) 支柱二：RL算法与架构 (RL & Architecture) (10 🔗1) 支柱八：物理动画 (Physics-based Animation) (1 🔗1) 支柱七：动作重定向 (Motion Retargeting) (1 🔗1) 支柱九：具身大模型 (Embodied Foundation Models) (1)

🔬 支柱一：机器人控制 (Robot Control) (35 篇)

#题目一句话要点标签🔗⭐
1 Mimetic Embodiment Transfer of Human-Humanoid Interaction from In-the-Wild Videos 提出HumiTransfer框架以解决人类与类人机器人自然交互问题 humanoid humanoid robot Unitree
2 VLA-Stem: A Lightweight Safety Mechanism for Vision-Language-Action Models via Reinforcement Learning 提出VLA-Stem以解决动态障碍物下的机器人安全控制问题 manipulation reinforcement learning curriculum learning
3 Trace-WAM: Learning to Act via 3D Motion Prediction for World Action Models 提出Trace-WAM以通过3D运动预测提升机器人控制能力 manipulation world action model world action models
4 Whole-Body Humanoid Control via Stochastic Motion Tokenization 提出随机运动标记化方法以解决人形机器人全身控制问题 quadruped humanoid humanoid control
5 HC2Robot: Hyper-Connection Transformers with Predicted Foresight for Multi-Terrain Humanoid Parkour 提出HC2Robot以解决多地形人形机器人跑酷问题 humanoid humanoid locomotion locomotion
6 ActiveAlign: Aligning Egocentric Manipulation via Active Perception 提出ActiveAlign以解决真实与模拟环境对齐问题 manipulation sim-to-real sim2real
7 RynnWorld-Latent: A Cross-Embodiment World Model Conditioned on Universal Latent Actions 提出RynnWorld-Latent以解决跨体态世界模型学习问题 teleoperation world model world models
8 Ego-CEVR: Learning Cross-Embodiment Visual Representations from Human Egocentric Videos 提出Ego-CEVR框架以解决机器人操作能力学习问题 manipulation representation learning egocentric
9 SWIM: Learning What Matters from Behaviors for JEPA World Models 提出SWIM以解决JEPAs在任务相关状态信息捕捉中的不足 manipulation world model world models
10 A fully articulated humanoid dexterous hand with human-level joint structure enabled by linkage-based actuation 提出Torch-Hand以解决人形灵巧手的运动复杂性与控制简易性问题 humanoid manipulation dexterous hand
11 REACH: An RL-based Hierarchical Humanoid Crawling Framework for Adaptable Confined Space Traversal 提出REACH框架以解决人形机器人在受限空间中的适应性爬行问题 humanoid humanoid locomotion bipedal ✅
12 ThinkWorld: World Imagination Reinforces Chain-of-Dynamics Reasoning 提出ThinkWorld框架以解决VLA模型泛化能力不足问题 manipulation reinforcement learning human-to-robot
13 Tokens of Motion: Primitive Motion Learning for Human-Humanoid Skill Transfer 提出自适应失败信息学习以解决机器人操作中的失败问题 humanoid manipulation policy learning
14 GAWM: A Compact Latent World-Action Model for Multi-Task Robot Manipulation 提出GAWM以解决多任务机器人操控中的高成本与延迟问题 manipulation dual-arm policy learning ✅
15 CARD: Contact-Anchored Retargeting for Dexterous Manipulation 提出CARD框架以解决机器人灵巧操作中的接触力保持问题 manipulation dexterous manipulation teleoperation
16 A Trainable CPG Framework for Quadruped Locomotion 提出可训练的CPG框架以优化四足动物运动控制 quadruped locomotion Unitree
17 GRAFT: Geometry-Aware Residual Adaptation for Foundation Trackers 提出GRAFT框架以解决地形适应性不足问题 humanoid locomotion teleoperation
18 TOPS: Reshaping Flow Source and Transport for Multi-Task Robot Manipulation 提出TOPS以解决多任务机器人操控中的流源与传输问题 manipulation flow matching multimodal
19 RoboSTD: Zero-Shot Single-to-Dual Transfer via Sagittal Mirroring for Bimanual Learning 提出RoboSTD以解决单臂到双臂转移问题 manipulation bi-manual policy learning ✅
20 Keeping Contact: Task-Adaptive Friction Margins for Bimanual Non-Prehensile Manipulation 提出任务自适应摩擦边际以解决双手非抓取操作中的接触保持问题 manipulation bi-manual
21 On Shaky Ground: Characterising Non-Rigid Base Effects in Mobile Manipulation 研究非刚性底座效应以优化移动机器人操作性能 manipulation mobile manipulation
22 Remember Where You Walk: Learning Spatial Memory for End-to-End Legged Robot Exploration 提出基于Hadamard记忆的LSMM以解决机器人导航中的空间记忆问题 legged robot humanoid reinforcement learning
23 When Black-Box Robot Policies Fail: Online Sensor Diagnosis and Repair with a Transparent Surrogate 提出透明替代模型以解决机器人传感器故障问题 locomotion locomotion policy ANYmal
24 Which Hand Was Late? Diagnosis for Robot Failure Recovery in Bimanual Tool Use 提出基于视觉-语言模型的双手工具使用故障诊断方法 manipulation bi-manual
25 EgoThumb: A Vision-Driven Supernumerary Robotic Thumb for Autonomous Grasp and Holding Assistance 提出EgoThumb以解决超数指机器人在复杂任务中的局限性 manipulation bi-manual imitation learning
26 Kinematic Constraint-Referenced Evaluation for Articulated Object Manipulation 提出约束参考策略以优化关节物体操控评估 manipulation behavior cloning
27 RoboAuto: Autonomous Robot Program Search through Multi-Level Physical Understanding 提出RoboAuto以解决机器人程序搜索中的物理理解问题 manipulation affordance Code as Policies
28 Camera-to-Quadruped Robot Pose Estimation from Sequential Images 提出QuadPose以解决四足机器人姿态估计问题 quadruped
29 OpenJaws: An Open Hardware UMI Framework with Stereo Localization for Generalist Manipulation 提出OpenJaws框架以解决视觉定位精度不足问题 manipulation
30 A generalist policy for muscle-driven control across individual musculoskeletal models 提出MyoAcross框架以解决肌肉驱动控制模型间政策重用问题 motion tracking policy learning
31 Pushing Dexterity and Safety for In-Hand Object Pose Reaching That Works Remarkably Well 提出一种新方法以提升多指机器人手的灵巧性与安全性 manipulation dexterous manipulation ✅
32 A Bi-level Co-design Framework for Lunar Monoped with 2D Jump and Landing Control 提出双层协同设计框架以优化月球单足机器人跳跃控制 legged robot trajectory optimization
33 When, Where, and What to Correct: Human-Guided Adaptive Residual Correction for Robot Policies 提出自适应残差修正以解决机器人精确操作问题 manipulation reinforcement learning
34 Learning Quality-Controllable Robot Policies with Iterative Hindsight Relabeling 提出质量可控的机器人策略以解决任务执行质量问题 manipulation
35 From Biological Curves to Soft Modular Robots: Learning-Based Inverse Design with Uncertainty-Guided Refinement 提出基于学习的逆向设计方法以解决软机器人形状重构问题 sim-to-real

🔬 支柱二：RL算法与架构 (RL & Architecture) (10 篇)

#题目一句话要点标签🔗⭐
36 PDRVLA: A Vision-Language-Action Model with Physical Dynamics Representations 提出PDRVLA以解决视觉-语言-动作模型对物理动态理解不足的问题 world model world models optical flow
37 World Dreamer v5-lite: A Multi-Robot World Action Model with One Dreamer 提出World Dreamer v5-lite以解决多机器人控制中的视角整合问题 world action model world action models dreamer
38 How Much Can Best-of-N Help a World Action Model? Measuring Selectable Headroom and Realizing It in Closed Loop 提出去噪反事实协议以提高世界动作模型的选择性表现 world action model world action models
39 Learning Visuomotor Quadrotor Flight Through Dynamic Narrow Gaps via Differentiable Simulation 提出可微仿真方法以解决四旋翼穿越动态狭窄间隙问题 distillation differentiable simulation
40 Rendering-Free Event Frame Simulation for Fast Reinforcement Learning of Autonomous Robot Navigation 提出无渲染事件帧模拟以加速自主机器人导航的强化学习 reinforcement learning traversability ✅
41 From 12 to 200 Demonstrations: How ACT, Diffusion Policy, SmolVLA and $π_{0.5}$ Scale on a Low-Cost Arm 通过多样化示例提升低成本机械臂的模仿学习效果 imitation learning diffusion policy
42 Predict the Swing, Match the Speed: Robotic Pipe Stabbing With Learned Stereo Vision and Residual Reinforcement Learning 提出StabBot以解决管道刺入过程中的速度匹配问题 reinforcement learning
43 Multi-Agent Reinforcement Learning of Co-Adaptive Hip Exoskeleton Control (MARL-Exo) for Metabolic Energy Reduction 提出MARL-Exo以降低步态行走的代谢能耗问题 reinforcement learning
44 TAPESTRY: A Single Backbone for Video Generation and Robot Control 提出TAPESTRY以解决视频生成与机器人控制的高成本问题 world action model world action models transformer policy
45 Learning Side-by-Side Navigation for Personal Mobility Vehicles from Human Drivers 提出一种新方法实现个人移动车辆的并行导航 reinforcement learning inverse reinforcement learning

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

#题目一句话要点标签🔗⭐
46 Retrieve to Act: Motion Primitive Graph Retrieval for Robust VLA Execution 提出R2A框架以解决视觉-语言-动作执行中的OOD鲁棒性问题 spatiotemporal vision-language-action VLA ✅

🔬 支柱七：动作重定向 (Motion Retargeting) (1 篇)

#题目一句话要点标签🔗⭐
47 RESPLE2: Inertial-Informed Recursive Non-Uniform Spline Estimation for Multi-Sensor Odometry with Global Refinement 提出RESPLE2以解决多传感器里程计精度不足问题 motion representation TAMP ✅

🔬 支柱九：具身大模型 (Embodied Foundation Models) (1 篇)

#题目一句话要点标签🔗⭐
48 Making Self-Driving Laboratories Auditable:A Unified Framework for Hybrid Actions,Measurements, and Safe Execution 提出可审计的自驾实验室框架以解决执行接口挑战 vision-language-action

⬅️ 返回 cs.RO 首页 · 🏠 返回主页