cs.RO（2026-03-10）

📊 共 35 篇论文 | 🔗 6 篇有代码

🎯 兴趣领域导航

支柱一：机器人控制 (Robot Control) (28 🔗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) (2 🔗2) 支柱四：生成式动作 (Generative Motion) (1) 支柱二：RL算法与架构 (RL & Architecture) (1)

🔬 支柱一：机器人控制 (Robot Control) (28 篇)

#题目一句话要点标签🔗⭐
1 Kinodynamic Motion Retargeting for Humanoid Locomotion via Multi-Contact Whole-Body Trajectory Optimization 提出KDMR框架，通过多接触全身轨迹优化实现拟人运动重定向 humanoid humanoid locomotion locomotion
2 ZeroWBC: Learning Natural Visuomotor Humanoid Control Directly from Human Egocentric Video ZeroWBC：直接从人类第一视角视频学习自然的人形机器人全身控制 humanoid humanoid robot humanoid control
3 TiPToP: A Modular Open-Vocabulary Planning System for Robotic Manipulation TiPToP：用于机器人操作的模块化开放词汇规划系统 manipulation open-vocabulary open vocabulary
4 DexHiL: A Human-in-the-Loop Framework for Vision-Language-Action Model Post-Training in Dexterous Manipulation DexHiL：用于灵巧操作中VLA模型人机协同后训练框架 manipulation dexterous hand dexterous manipulation ✅
5 SCDP: Learning Humanoid Locomotion from Partial Observations via Mixed-Observation Distillation SCDP：通过混合观测蒸馏学习部分观测下的人形机器人运动控制 humanoid humanoid locomotion locomotion
6 NS-VLA: Towards Neuro-Symbolic Vision-Language-Action Models 提出神经符号视觉-语言-动作模型，提升机器人操作任务中的数据效率和泛化性 manipulation reinforcement learning vision-language-action
7 Cutting the Cord: System Architecture for Low-Cost, GPU-Accelerated Bimanual Mobile Manipulation 提出一种低成本、GPU加速的双臂移动操作机器人系统架构。 manipulation mobile manipulation bi-manual
8 See, Plan, Rewind: Progress-Aware Vision-Language-Action Models for Robust Robotic Manipulation 提出SPR框架，通过显式里程碑实现鲁棒的机器人操作 manipulation vision-language-action OpenVLA
9 Robust Spatiotemporal Motion Planning for Multi-Agent Autonomous Racing via Topological Gap Identification and Accelerated MPC 提出基于拓扑间隙识别与加速MPC的多智能体自主赛车鲁棒时空运动规划方法 MPC motion planning spatiotemporal
10 Beyond Short-Horizon: VQ-Memory for Robust Long-Horizon Manipulation in Non-Markovian Simulation Benchmarks 提出VQ-Memory，用于提升非马尔可夫环境中长时程机械臂操作的鲁棒性。 manipulation diffusion policy VQ-VAE
11 Emerging Extrinsic Dexterity in Cluttered Scenes via Dynamics-aware Policy Learning 提出动力学感知策略学习框架，解决复杂场景下的外在灵巧操作问题 manipulation sim-to-real teleoperation
12 SEA-Nav: Efficient Policy Learning for Safe and Agile Quadruped Navigation in Cluttered Environments 提出SEA-Nav框架，解决复杂环境中四足机器人安全高效导航问题 quadruped locomotion reinforcement learning
13 MO-Playground: Massively Parallelized Multi-Objective Reinforcement Learning for Robotics 提出MORLAX算法和MO-Playground环境，加速多目标机器人强化学习。 humanoid humanoid robot locomotion
14 Towards Terrain-Aware Safe Locomotion for Quadrupedal Robots Using Proprioceptive Sensing 针对四足机器人，提出基于本体感觉的安全地形感知运动方法 quadruped locomotion
15 Trajectory Optimization for Self-Wrap-Aware Cable-Towed Planar Object Manipulation under Implicit Tension Constraints 提出自缠绕感知的绳索牵引平面物体轨迹优化方法，解决隐式张力约束下的操作问题。 manipulation trajectory optimization
16 Embodied Human Simulation for Quantitative Design and Analysis of Interactive Robotics 提出基于具身人类仿真的交互机器人量化设计与分析框架 humanoid humanoid robot reinforcement learning
17 From Flow to One Step: Real-Time Multi-Modal Trajectory Policies via Implicit Maximum Likelihood Estimation-based Distribution Distillation 提出基于隐式最大似然估计的分布蒸馏方法，加速多模态轨迹策略，实现机器人实时操作。 manipulation flow matching distillation
18 CORAL: Scalable Multi-Task Robot Learning via LoRA Experts CORAL：基于LoRA专家网络的可扩展多任务机器人学习框架 dual-arm vision-language-action VLA ✅
19 TRIP-Bag: A Portable Teleoperation System for Plug-and-Play Robotic Arms and Leaders TRIP-Bag：一种便携式遥操作系统，用于即插即用型机械臂和示教器 manipulation teleoperation
20 Robotic Scene Cloning:Advancing Zero-Shot Robotic Scene Adaptation in Manipulation via Visual Prompt Editing 提出Robotic Scene Cloning，通过视觉提示编辑实现机器人零样本场景适应 manipulation
21 MuxGel: Simultaneous Dual-Modal Visuo-Tactile Sensing via Spatially Multiplexing and Deep Reconstruction MuxGel：通过空间复用和深度重建实现同步双模态视觉触觉传感 manipulation sim-to-real ✅
22 Stein Variational Ergodic Surface Coverage with SE(3) Constraints 提出基于SE(3)约束的Stein变分遍历表面覆盖方法，用于机器人操作复杂3D表面 manipulation trajectory optimization
23 Walking on Rough Terrain with Any Number of Legs 提出一种轻量级多足机器人控制架构，提升复杂地形适应性 legged robot locomotion
24 Provably Safe Trajectory Generation for Manipulators Under Motion and Environmental Uncertainties 提出基于RM-DeSKO和SOS验证的风险约束机械臂轨迹生成方法 sim-to-real motion planning
25 ReTac-ACT: A State-Gated Vision-Tactile Fusion Transformer for Precision Assembly 提出ReTac-ACT，解决末端执行器遮挡下高精度装配中的视觉失效问题。 manipulation imitation learning
26 TIMID: Time-Dependent Mistake Detection in Videos of Robot Executions TIMID：用于检测机器人执行视频中时序依赖错误的视频异常检测架构 sim-to-real ✅
27 Caterpillar-Inspired Spring-Based Compressive Continuum Robot for Bristle-based Exploration 提出一种受毛毛虫启发的弹簧压缩连续体机器人，用于基于刷毛的管道空间探索 locomotion
28 Quality over Quantity: Demonstration Curation via Influence Functions for Data-Centric Robot Learning QoQ：基于影响函数的数据驱动机器人学习高质量示教数据筛选方法 teleoperation

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

#题目一句话要点标签🔗⭐
29 SPAN-Nav: Generalized Spatial Awareness for Versatile Vision-Language Navigation SPAN-Nav：通过通用空间感知实现多功能视觉-语言导航 VLN foundation model chain-of-thought
30 Let's Reward Step-by-Step: Step-Aware Contrastive Alignment for Vision-Language Navigation in Continuous Environments 提出SACA框架，解决连续环境VLN中奖励稀疏和误差累积问题 VLN large language model
31 PM-Nav: Priori-Map Guided Embodied Navigation in Functional Buildings 提出PM-Nav，利用先验地图引导功能建筑中的具身导航 chain-of-thought

🔬 支柱三：空间感知与语义 (Perception & Semantics) (2 篇)

#题目一句话要点标签🔗⭐
32 BEACON: Language-Conditioned Navigation Affordance Prediction under Occlusion BEACON：解决遮挡下语言条件导航的Affordance预测问题 open-vocabulary open vocabulary affordance ✅
33 STONE Dataset: A Scalable Multi-Modal Surround-View 3D Traversability Dataset for Off-Road Robot Navigation STONE数据集：用于越野机器人导航的可扩展多模态环视3D可通行性数据集 traversability ✅

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
34 StyleVLA: Driving Style-Aware Vision Language Action Model for Autonomous Driving StyleVLA：面向自动驾驶的风格感知视觉语言动作模型 physically plausible first-person view vision-language-action

🔬 支柱二：RL算法与架构 (RL & Architecture) (1 篇)

#题目一句话要点标签🔗⭐
35 Latent World Models for Automated Driving: A Unified Taxonomy, Evaluation Framework, and Open Challenges 提出统一的自动驾驶潜在世界模型框架，并提供评估标准与未来研究方向。 world model vision-language-action VLA

⬅️ 返回 cs.RO 首页 · 🏠 返回主页