cs.RO（2026-03-05）

📊 共 42 篇论文 | 🔗 6 篇有代码

🎯 兴趣领域导航

支柱一：机器人控制 (Robot Control) (30 🔗4) 支柱二：RL算法与架构 (RL & Architecture) (4) 支柱九：具身大模型 (Embodied Foundation Models) (3) 支柱三：空间感知与语义 (Perception & Semantics) (3 🔗2) 支柱四：生成式动作 (Generative Motion) (1) 支柱五：交互与反应 (Interaction & Reaction) (1)

🔬 支柱一：机器人控制 (Robot Control) (30 篇)

#题目一句话要点标签🔗⭐
1 cuRoboV2: Dynamics-Aware Motion Generation with Depth-Fused Distance Fields for High-DoF Robots cuRoboV2：基于深度融合距离场和动力学的多自由度机器人运动生成框架 humanoid locomotion locomotion policy
2 PhysiFlow: Physics-Aware Humanoid Whole-Body VLA via Multi-Brain Latent Flow Matching and Robust Tracking PhysiFlow：基于物理感知的多脑潜在流匹配人形机器人全身VLA控制 humanoid humanoid robot whole-body control
3 Omni-Manip: Beyond-FOV Large-Workspace Humanoid Manipulation with Omnidirectional 3D Perception Omni-Manip：基于全向3D感知的类人机器人大范围操作 humanoid humanoid robot manipulation
4 Iterative On-Policy Refinement of Hierarchical Diffusion Policies for Language-Conditioned Manipulation HD-ExpIt：通过迭代式On-Policy优化分层扩散策略，解决语言条件下的机械臂操作任务。 manipulation diffusion policy language conditioned
5 Critic in the Loop: A Tri-System VLA Framework for Robust Long-Horizon Manipulation 提出Critic in the Loop框架，解决长时程操作中VLM推理延迟与VLA语义不足的矛盾 manipulation vision-language-action VLA
6 GaussTwin: Unified Simulation and Correction with Gaussian Splatting for Robotic Digital Twins GaussTwin：利用高斯溅射实现机器人数字孪生的统一仿真与校正 manipulation model predictive control gaussian splatting
7 Lifelong Language-Conditioned Robotic Manipulation Learning 提出SkillsCrafter框架，解决语言条件机器人操作中终身学习的灾难性遗忘问题。 manipulation language conditioned
8 SeedPolicy: Horizon Scaling via Self-Evolving Diffusion Policy for Robot Manipulation SeedPolicy：通过自进化扩散策略实现机器人操作的水平扩展 manipulation imitation learning diffusion policy ✅
9 VPWEM: Non-Markovian Visuomotor Policy with Working and Episodic Memory VPWEM：利用工作记忆和情景记忆的非马尔可夫视觉运动策略，提升长时依赖任务性能。 manipulation mobile manipulation imitation learning ✅
10 UltraDexGrasp: Learning Universal Dexterous Grasping for Bimanual Robots with Synthetic Data UltraDexGrasp：面向双臂机器人的通用灵巧抓取学习框架，基于合成数据。 manipulation bi-manual sim-to-real ✅
11 Task-Relevant and Irrelevant Region-Aware Augmentation for Generalizable Vision-Based Imitation Learning in Agricultural Manipulation 提出DRAIL框架，通过区域感知增强提升农业操作中视觉模仿学习的泛化性 manipulation imitation learning diffusion policy
12 Residual RL--MPC for Robust Microrobotic Cell Pushing Under Time-Varying Flow 提出基于残差强化学习的MPC方法，用于时变流场下稳健的微型机器人细胞推运。 manipulation MPC SAC
13 Hyperbolic Multiview Pretraining for Robotic Manipulation HyperMVP：提出基于双曲空间的多视角预训练框架，提升机器人操作任务性能。 manipulation masked autoencoder visual pre-training
14 Direct Contact-Tolerant Motion Planning With Vision Language Models 提出基于视觉语言模型的直接接触容忍运动规划方法，提升复杂环境导航能力 motion planning contact-aware ✅
15 SPIRIT: Perceptive Shared Autonomy for Robust Robotic Manipulation under Deep Learning Uncertainty 提出基于深度学习不确定性的感知共享自主控制方法，用于提升机器人操作的鲁棒性。 manipulation teleoperation
16 LEGS-POMDP: Language and Gesture-Guided Object Search in Partially Observable Environments 提出LEGS-POMDP以解决部分可观测环境中的物体搜索问题 quadruped foundation model multimodal
17 Observer Design for Augmented Reality-based Teleoperation of Soft Robots 提出基于增强现实的软体机器人遥操作观测器设计方案 teleoperation
18 Designing and Validating a Self-Aligning Tool Changer for Modular Reconfigurable Manipulation Robots 设计自对准工具更换器，解决模块化重构机器人工具自动更换难题 manipulation
19 Safe-SAGE: Social-Semantic Adaptive Guidance for Safe Engagement through Laplace-Modulated Poisson Safety Functions Safe-SAGE：基于泊松安全函数和拉普拉斯引导场的语义自适应安全控制框架 legged robot model predictive control
20 Accelerating Sampling-Based Control via Learned Linear Koopman Dynamics 提出基于学习线性Koopman动态的加速采样控制方法，提升复杂非线性系统控制效率。 quadruped
21 Beyond the Patch: Exploring Vulnerabilities of Visuomotor Policies via Viewpoint-Consistent 3D Adversarial Object 提出基于可微渲染的视角一致3D对抗物体，提升机器人视觉策略的鲁棒性 manipulation
22 Gait Generation Balancing Joint Load and Mobility for Legged Modular Robots with Easily Detachable Joints 针对易拆卸关节的模块化腿式机器人，提出平衡关节负载和移动性的步态生成方法 locomotion
23 Safe-Night VLA: Seeing the Unseen via Thermal-Perceptive Vision-Language-Action Models for Safety-Critical Manipulation 提出Safe-Night VLA，用于夜间热感知操作的安全关键型视觉-语言-动作模型 manipulation vision-language-action VLA
24 OmniDP: Beyond-FOV Large-Workspace Humanoid Manipulation with Omnidirectional 3D Perception OmniDP：基于全向3D感知的超视场人形机器人操作 humanoid humanoid robot manipulation
25 EmboAlign: Aligning Video Generation with Compositional Constraints for Zero-Shot Manipulation EmboAlign：通过组合约束对齐视频生成，实现零样本操作 manipulation trajectory optimization depth estimation
26 PRISM: Personalized Refinement of Imitation Skills for Manipulation via Human Instructions PRISM：通过人类指令个性化改进操作机器人的模仿技能 manipulation reinforcement learning imitation learning
27 Contact-Grounded Policy: Dexterous Visuotactile Policy with Generative Contact Grounding Contact-Grounded Policy：通过生成式接触感知实现灵巧操作 manipulation dexterous manipulation teleoperation
28 Safe-SAGE: Social-Semantic Adaptive Guidance for Safe Engagement through Laplace-Modulated Poisson Safety Functions Safe-SAGE：基于拉普拉斯调制泊松安全函数的社交语义自适应引导安全交互 legged robot model predictive control
29 TEGA: A Tactile-Enhanced Grasping Assistant for Assistive Robotics via Sensor Fusion and Closed-Loop Haptic Feedback TEGA：触觉增强的抓取辅助系统，通过传感器融合和闭环触觉反馈实现 manipulation teleoperation
30 Task Parameter Extrapolation via Learning Inverse Tasks from Forward Demonstrations 提出基于逆任务学习的任务参数外推方法，提升机器人泛化操作能力 manipulation imitation learning

🔬 支柱二：RL算法与架构 (RL & Architecture) (4 篇)

#题目一句话要点标签🔗⭐
31 Latent Policy Steering through One-Step Flow Policies 提出LPS，通过单步流策略实现离线强化学习中的高效潜在策略引导 reinforcement learning offline RL offline reinforcement learning
32 LLM-Guided Decentralized Exploration with Self-Organizing Robot Teams 提出基于LLM引导的去中心化探索方法，实现机器人团队的自主自组织 reinforcement learning deep reinforcement learning large language model
33 Relational Semantic Reasoning on 3D Scene Graphs for Open World Interactive Object Search 提出SCOUT，利用场景图关系推理实现开放世界交互式物体搜索 distillation open-vocabulary open vocabulary
34 TransMASK: Masked State Representation through Learned Transformation TransMASK：通过学习变换的掩码状态表示，提升机器人策略泛化能力 imitation learning diffusion policy

🔬 支柱九：具身大模型 (Embodied Foundation Models) (3 篇)

#题目一句话要点标签🔗⭐
35 Observing and Controlling Features in Vision-Language-Action Models 提出特征可观测性和可控性方法，实现对Vision-Language-Action模型行为的精准调控。 vision-language-action VLA large language model
36 OpenFrontier: General Navigation with Visual-Language Grounded Frontiers 提出OpenFrontier以解决开放世界导航问题 VLA VLN
37 On the Strengths and Weaknesses of Data for Open-set Embodied Assistance 提出开放集纠正辅助框架，研究数据多样性对具身智能模型泛化能力的影响 foundation model multimodal

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

#题目一句话要点标签🔗⭐
38 VinePT-Map: Pole-Trunk Semantic Mapping for Resilient Autonomous Robotics in Vineyards VinePT-Map：面向葡萄园的杆-干语义地图，提升自主机器人的长期鲁棒性 semantic mapping semantic map
39 AIM-SLAM: Dense Monocular SLAM via Adaptive and Informative Multi-View Keyframe Prioritization with Foundation Model AIM-SLAM：利用自适应多视角关键帧和几何基础模型实现稠密单目SLAM VGGT foundation model ✅
40 AIM-SLAM: Dense Monocular SLAM via Adaptive and Informative Multi-View Keyframe Prioritization with Foundation Model AIM-SLAM：利用自适应信息多视角关键帧优先级的稠密单目SLAM VGGT foundation model ✅

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
41 CT-Enabled Patient-Specific Simulation and Contact-Aware Robotic Planning for Cochlear Implantation 提出CT引导的患者特异性仿真与接触感知机器人规划，用于耳蜗植入 contact-aware

🔬 支柱五：交互与反应 (Interaction & Reaction) (1 篇)

#题目一句话要点标签🔗⭐
42 RFM-HRI : A Multimodal Dataset of Medical Robot Failure, User Reaction and Recovery Preferences for Item Retrieval Tasks RFM-HRI：针对医疗机器人故障，用户反应及恢复偏好的多模态数据集 dyadic interaction multimodal

⬅️ 返回 cs.RO 首页 · 🏠 返回主页