cs.RO（2026-10-13）

📊 共 70 篇论文 | 🔗 15 篇有代码

🎯 兴趣领域导航

支柱一：机器人控制 (Robot Control) (44 🔗11) 支柱二：RL算法与架构 (RL & Architecture) (15 🔗3) 支柱九：具身大模型 (Embodied Foundation Models) (7 🔗1) 支柱三：空间感知与语义 (Perception & Semantics) (3) 支柱四：生成式动作 (Generative Motion) (1)

🔬 支柱一：机器人控制 (Robot Control) (44 篇)

#题目一句话要点标签🔗⭐
1 Learning Robot Policies for Whole-Body Mobility Assistance via Dynamic Human Motion Generation 提出基于动态人类动作生成的机器人全身移动辅助策略 sim-to-real diffusion policy motion generation ✅
2 TRACE: Efficient Transfer Learning for Neural Whole-Body Control Across Humanoid Embodiments 提出TRACE框架以解决类人机器人控制的跨体型适应问题 humanoid humanoid control whole-body control
3 YUBI-Real2Sim: 3DGS Scene Verification for Reliable Policy Evaluation 提出YUBI-Real2Sim以解决真实机器人政策评估中的场景重建问题 bi-manual real2sim 3D gaussian splatting
4 EgoExplorer: A Vision-Motion-Touch Observatory for Human Loco-Manipulation In Situ 提出EgoExplorer以解决人类运动与操控数据收集不足问题 humanoid whole-body control locomotion
5 DR-VLA: Learning Human-Like Quadruped Navigation from Egocentric Web Videos via Domain-Invariant Scene Representations 提出DR-VLA以解决四足机器人导航中的场景理解问题 quadruped flow matching world model
6 Breaking the Average: Overcoming Convergent Defaults in Robotic Manipulation via Diagnostic Chain-of-Thought and Structured Guidance 提出新框架以解决机器人操作中的收敛默认问题 manipulation behavior cloning vision-language-action
7 EgoCPT: Learning Dexterous Manipulation from a Single Ego-Centric Video with Contact-Aware Optimization 提出EgoCPT以解决从人类视频学习灵巧操作的挑战 manipulation dexterous hand dexterous manipulation
8 Cross-Embodiment Affordance Transfer for Robotic Manipulation 提出跨体现 affordance 转移框架以解决机器人操作中的人机不匹配问题 manipulation affordance cross-embodiment
9 WAVE-0: Bridging VLA Policies and WAMs with Lightweight Latent Future Prediction 提出WAVE-0框架以解决VLA与WAM间的整合问题 manipulation world model world models
10 GreenWAM: A World Foundation Model for Robotics 提出GreenWAM以解决机器人世界模型与策略选择的双重问题 humanoid bi-manual world model
11 EgoMatrix: Physically Faithful Dexterous Robot Demonstrations from Egocentric Human Videos 提出EgoMatrix以解决从第一人称视频获取机器人演示数据的问题 manipulation dexterous manipulation bi-manual ✅
12 Walk Like This: Stylized Locomotion for Physics-Based Characters through Example-Driven Diffusion 提出基于示例驱动扩散的风格化步态控制方法 legged locomotion humanoid locomotion
13 ForkCast: Predictive Recovery for Vision-Language-Action Policies via Counterfactual Futures 提出ForkCast框架以解决视觉-语言-动作策略中的恢复问题 manipulation world model world models
14 Interaction Latents as Actions for Contact-Rich Manipulation 提出ILA方法以解决接触丰富的机器人操作问题 manipulation flow matching world action model ✅
15 I3L: Iterative Interactive Imitation Learning for Whole-Body Mobile Manipulation 提出I3L以解决长时间全身移动操控的可靠性问题 manipulation mobile manipulation imitation learning ✅
16 EqC: Equilibrium Energy Control for Introspective Lifelong Robot Learning 提出EqC框架以解决机器人自主学习中的行为一致性问题 manipulation bi-manual bimanual manipulation
17 Let the Mind Conceive, Let the Hand Achieve: Following Multimodal Instructions for Long-Horizon Robotic Manipulation 提出Orchestra以解决长时间机器人操作中的多步骤指令问题 manipulation multimodal ✅
18 RoboSimStudio: An Interactive Toolkit for Rapid Real2Sim of Robot Manipulation Tasks 提出RoboSimStudio以解决机器人操控任务的真实与模拟间的物理建模问题 manipulation sim2real real2sim
19 CoSMO: Conformal Semantic Mapping for Open-World Safe Robot Navigation 提出CoSMO框架以解决开放世界安全机器人导航问题 quadruped semantic mapping semantic map ✅
20 DexStitch: Learning and Composing Dexterous Skills from Human Videos via Contact-Aware Robotic Inpainting 提出DexStitch以解决机器人技能学习中的视频转移问题 manipulation dexterous hand teleoperation
21 Acting Through Registers 提出RegisterWAM以解决WAM中空间异常值问题 manipulation flow matching world model
22 ForceWAM: Decoupling Future and Response Learning from Demo and Play Data for Contact-Rich Manipulation 提出ForceWAM以解决接触丰富任务中的动态学习问题 manipulation world model world models ✅
23 WorldOracle: Benchmarking Action-Conditioned World Models against Real-World Executions 提出WorldOracle基准以评估动作条件世界模型的真实执行能力 manipulation world model world models
24 LynxScenes: A Large-Scale Real-World Benchmark for Outdoor Wheel-Legged Robot Perception, Navigation, Imagination and Reconstruction 提出LynxScenes数据集以解决户外轮腿机器人感知与导航问题 legged robot world model world models
25 From Pixels to Executable Physics: Behavior-Grounded World Models for Embodied Action Consequence Prediction 提出DR-WM以解决四足机器人动作结果预测问题 quadruped world model world models
26 Box-WAM: Box Trajectories as Compact State Representations for World Action Models 提出Box-WAM以解决视频预测计算成本高的问题 manipulation world action model world action models
27 CALM: Decentralized Control of Cooperative Aerial Manipulation with Multi-Agent Reinforcement Learning 提出CALM以解决多无人机协作操控中的挑战 manipulation reinforcement learning ✅
28 RESTA: Time-Scaling Augmentation for Data-Efficient Robot Imitation Learning 提出RESTA以提高机器人模仿学习的数据效率 manipulation bi-manual bimanual manipulation
29 Benchmarking Interaction-Oriented Aerial Mobility across Heterogeneous Environments 提出CrossFly基准以解决异构环境下无人机交互移动问题 manipulation mobile manipulation vision-language-action
30 LAG-IO: Multi-IMU State Estimation for Legged Robots Using Factor Graphs 提出LAG-IO以解决腿部机器人状态估计问题 quadruped legged robot locomotion
31 PACT: Posture-Aligned Co-Design Technique for Hand Configuration and Retargeting in Multi-Finger Imitation Learning 提出PACT以优化多指模仿学习中的手部配置与重定向 teleoperation imitation learning human-to-robot
32 Beyond Safe Execution: Safe Online Reinforcement Learning for Autonomous Driving with Value Calibration and Adaptive Guidance 提出SOAR-Flow框架以解决自主驾驶中的安全在线强化学习问题 MPC model predictive control reinforcement learning ✅
33 RoboFQE: Rollout-Free Evaluation of Visual Robot Policies from Offline Trajectories 提出RoboFQE以解决机器人策略评估中的回滚需求问题 manipulation world model world models
34 Locomotion with Online Awareness of Dynamics for Precision Footholds 提出LOAD框架以解决未知负载下的精确步态问题 locomotion height map
35 Beyond the End-Effector: Learning Whole-Arm Contact Manipulation from Human Demonstrations 提出Whole-Arm Contact Learning以解决机器人接触操控问题 manipulation teleoperation
36 LADDER: A Long-Horizon Robot Manipulation Benchmark 提出LADDER基准以解决长时间机器人操作任务的挑战 manipulation
37 ReEnvision: Future-Guided Selective Action Repair for Long-Horizon Manipulation 提出ReEnvision以解决长时间机器人操作中的选择性动作修复问题 manipulation
38 UniW(A)M: A Unified Mixture-of-Experts Framework for World and World-Action Models 提出UniW(A)M框架以解决世界模型与动作模型的知识孤岛问题 manipulation world model world models
39 ForeTac: Predictive Contact Guidance for Generative Robot Policies 提出ForeTac以解决接触丰富操作中的反馈延迟问题 manipulation diffusion policy flow matching ✅
40 Contact Trajectory Prompting: In-Context Transfer of Contact-Rich Behaviors from a Single Demonstration 提出接触轨迹提示以解决单次演示的接触丰富行为转移问题 manipulation ✅
41 Intent-guided Adaptive Stiffening for Visuomotor Policies under Unknown Resistance 提出意图引导自适应刚度以解决未知阻力下的操控问题 manipulation
42 GLST: Global-Local Spatio-Temporal Memory Decomposition Framework for Online Chained-Goal Navigation 提出GLST框架以解决在线链式目标导航问题 quadruped
43 Physics-Aware Object-Centric SE(3) Forecasting for Real-Time Proactive Dynamic Grasping 提出基于物理的物体中心SE(3)预测以解决动态抓取问题 reachability-aware
44 TacSpin: Tactile Tracking of In-Gripper Object Rotation on SO(3) via Axis-Magnitude Estimation 提出TacSpin以解决抓取物体旋转跟踪问题 manipulation

🔬 支柱二：RL算法与架构 (RL & Architecture) (15 篇)

#题目一句话要点标签🔗⭐
45 WoVR+: World Model as Reliable Simulator for Versatile VLA Policy Optimization with RL 提出WoVR+以解决VLA-RL中的幻觉问题 reinforcement learning world model world models ✅
46 ACCRUE: Continual Off-Policy Reinforcement Learning for Robot Foundation Models 提出ACCRUE以解决机器人基础模型的持续学习问题 reinforcement learning VLA foundation model
47 PrT-Flow: Probability-Tractable Flow Matching for Reinforcement Learning 提出PrT-Flow以解决强化学习中的概率密度计算问题 reinforcement learning SAC imitation learning ✅
48 Agentic Robot Creation through Generation, Verification, and Revision 提出Agentic机器人创建框架以解决高层用户意图转化问题 policy learning motion generation multimodal
49 The Human--VLA Communication Gap: What Humans Leave Implicit and What Policies Cannot Recover 提出人机通信模型以解决机器人执行任务中的信息与执行差距问题 policy learning VLA ✅
50 Grid or Nodes? Probing Object Memory in World Models Under Partial Observability 提出FloWM以解决机器人部分可观测环境中的物体记忆问题 world model world models
51 SkipWorld: Latent Skip Planning with Distilled World Models for Efficient Visual Navigation 提出SkipWorld以解决长时间规划效率低下问题 world model world models
52 Data-Free Continual Learning for World Action Model with Rectified Generative Replay 提出R²-WAM以解决机器人任务遗忘问题 world action model world action models
53 Augmenting Real-time Visualization in Minimally Invasive Surgery with Particle Filtering and Depth-Aware Occlusion Handling 提出实时可视化框架以解决微创手术中的遮挡问题 MAE depth estimation monocular depth
54 OccWAM: Occupancy World-Action Models with Joint Scene-to-Ego Flow Matching 提出OccWAM以解决现有占用预测与动作生成的耦合问题 flow matching world model world models
55 Training Generalist Navigation Policies via Action Self-Distillation 提出行动自蒸馏方法以解决机器人导航中的语言条件学习问题 distillation language conditioned
56 Steertuning Policies from a Few Human Videos 提出Steertuning以解决机器人学习任务适应性不足的问题 diffusion policy egocentric embodiment transfer
57 Collaborative Planning for UAV-UGV Teams under Operational Constraints via Masked Error-bounded Neuro-Symbolic Diffusion 提出MEND以解决UAV-UGV团队协作规划中的操作约束问题 diffusion policy traversability MDM
58 OneDemo: Unified Policy Learning and Data Generation from a Single Demonstration 提出OneDemo框架以解决单一示范数据生成与策略学习问题 reinforcement learning policy learning
59 LoopWAM: Closed-Loop Generalist Policy Improvement from Deployment Rollouts 提出LoopWAM以解决机器人政策在部署中的改进问题 world model world models

🔬 支柱九：具身大模型 (Embodied Foundation Models) (7 篇)

#题目一句话要点标签🔗⭐
60 What Matters in Quality-Aware VLA Pretraining: A Benchmark and Empirical Study 提出AXIS-Bench以优化视觉-语言-动作预训练中的示范质量利用 vision-language-action VLA ✅
61 ChunkAct: Action Chunking for Agentic Robotics 提出ChunkAct以提高代理机器人任务执行效率 vision-language-action VLA large language model
62 LIBERO-RIG: A Hardware-Setup Robustness Benchmark for Vision-Language-Action Models 提出LIBERO-RIG基准以评估视觉-语言-动作模型的硬件鲁棒性 vision-language-action
63 SEAR: Self-Evolving Agentic Robotics 提出自我进化的智能机器人框架以提升自主学习能力 vision-language-action multimodal Code as Policies
64 EmbodiedEvo: Self-improving Coding Agents for Physical Intelligence 提出EmbodiedEvo以解决物理智能中的编码代理自我改进问题 vision-language-action VLA
65 MemoryVLN: From Ego-Aware Memory Representation to Cross-Context Memory Reuse in Vision-Language Navigation 提出MemoryVLN以解决长轨迹导航中的记忆管理问题 VLN
66 Contractor Filters 提出承包商滤波器以解决非线性状态估计问题 multimodal

🔬 支柱三：空间感知与语义 (Perception & Semantics) (3 篇)

#题目一句话要点标签🔗⭐
67 Perception Beyond the Visible: Incompleteness-Aware Scene Understanding for Robot Autonomy 提出不完整性感知场景理解以解决机器人自主性问题 scene understanding
68 RTISSLAM: In-air Imaging Sonar SLAM using a Robust Pose Graph and Occupancy Grid Mapping 提出RTISSLAM以解决室内声纳SLAM中的反射干扰问题 occupancy grid
69 I2O2I-Nav: A Benchmark with a Modular Agent for Long-Horizon Indoor-to-Outdoor-to-Indoor Navigation 提出I2O2I-Nav以解决长距离室内外导航问题 3DGS

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
70 Code2Worlds-2: Physically Plausible Agent Harness for Embodied World Simulation 提出Code2Worlds-2以解决动态世界生成的物理语义缺失问题 physically plausible penetration large language model

⬅️ 返回 cs.RO 首页 · 🏠 返回主页