cs.RO（2026-03-18）

📊 共 20 篇论文 | 🔗 4 篇有代码

🎯 兴趣领域导航

支柱一：机器人控制 (Robot Control) (8 🔗1) 支柱三：空间感知与语义 (Perception & Semantics) (5 🔗2) 支柱九：具身大模型 (Embodied Foundation Models) (4) 支柱二：RL算法与架构 (RL & Architecture) (2 🔗1) 支柱六：视频提取与匹配 (Video Extraction) (1)

🔬 支柱一：机器人控制 (Robot Control) (8 篇)

#题目一句话要点标签🔗⭐
1 RoboForge: Physically Optimized Text-guided Whole-Body Locomotion for Humanoids RoboForge：面向人形机器人的物理优化文本引导全身运动框架 humanoid humanoid robot humanoid locomotion
2 ProbeFlow: Training-Free Adaptive Flow Matching for Vision-Language-Action Models ProbeFlow：面向VLA模型的免训练自适应Flow Matching加速框架 manipulation flow matching vision-language-action
3 DexViTac: Collecting Human Visuo-Tactile-Kinematic Demonstrations for Contact-Rich Dexterous Manipulation DexViTac：用于灵巧操作的视觉-触觉-运动学多模态数据采集系统 manipulation dexterous manipulation representation learning ✅
4 KineVLA: Towards Kinematics-Aware Vision-Language-Action Models with Bi-Level Action Decomposition KineVLA：提出一种双层动作分解的运动学感知视觉-语言-动作模型 manipulation vision-language-action VLA
5 VolumeDP: Modeling Volumetric Representation for Manipulation Policy Learning VolumeDP：通过建模体积表示提升操作策略学习性能 manipulation policy learning imitation learning
6 REAL: Robust Extreme Agility via Spatio-Temporal Policy Learning and Physics-Guided Filtering REAL：基于时空策略学习和物理引导滤波的鲁棒极限敏捷控制 quadruped parkour Unitree
7 EVA: Aligning Video World Models with Executable Robot Actions via Inverse Dynamics Rewards EVA：通过逆动力学奖励对齐视频世界模型与可执行机器人动作，解决执行鸿沟问题。 bi-manual reinforcement learning world model
8 Full Stack Navigation, Mapping, and Planning for the Lunar Autonomy Challenge 为月球自主挑战赛设计全栈导航、建图与规划系统，荣获第一名 motion planning visual odometry

🔬 支柱三：空间感知与语义 (Perception & Semantics) (5 篇)

#题目一句话要点标签🔗⭐
9 SafeLand: Safe Autonomous Landing in Unknown Environments with Bayesian Semantic Mapping SafeLand：基于贝叶斯语义地图的未知环境安全自主着陆系统 semantic mapping semantic map ✅
10 AgentVLN: Towards Agentic Vision-and-Language Navigation 提出AgentVLN，解决视觉语言导航中空间感知和表征不一致问题 metric depth VLN chain-of-thought ✅
11 P$^{3}$Nav: End-to-End Perception, Prediction and Planning for Vision-and-Language Navigation 提出P$^{3}$Nav，通过感知、预测和规划一体化框架提升视觉语言导航性能。 scene understanding semantic map VLN
12 Visual SLAM with DEM Anchoring for Lunar Surface Navigation 提出DEM约束的视觉SLAM，用于月球表面长距离导航 visual SLAM
13 Neural Radiance Maps for Extraterrestrial Navigation and Path Planning 提出基于NeRF的星外导航方法，解决行星探测车全局地图构建与路径规划问题 NeRF neural radiance field

🔬 支柱九：具身大模型 (Embodied Foundation Models) (4 篇)

#题目一句话要点标签🔗⭐
14 HeiSD: Hybrid Speculative Decoding for Embodied Vision-Language-Action Models with Kinematic Awareness HeiSD：基于运动学感知的具身视觉-语言-动作模型混合推测解码加速框架 vision-language-action VLA
15 Specification-Aware Distribution Shaping for Robotics Foundation Models 提出一种规范感知的分布调整框架，增强机器人基础模型对时序逻辑约束的满足能力 foundation model
16 FloorPlan-VLN: A New Paradigm for Floor Plan Guided Vision-Language Navigation 提出FloorPlan-VLN，利用语义平面图引导视觉-语言导航，提升空间推理能力。 VLN
17 AERR-Nav: Adaptive Exploration-Recovery-Reminiscing Strategy for Zero-Shot Object Navigation 提出AERR-Nav以解决多层环境中的零-shot物体导航问题 large language model multimodal

🔬 支柱二：RL算法与架构 (RL & Architecture) (2 篇)

#题目一句话要点标签🔗⭐
18 Generative Control as Optimization: Time Unconditional Flow Matching for Adaptive and Robust Robotic Control 提出GeCO：一种时间无条件Flow Matching框架，用于自适应和鲁棒的机器人控制。 imitation learning flow matching vision-language-action ✅
19 Interpreting Context-Aware Human Preferences for Multi-Objective Robot Navigation 提出一种融合VLM和LLM的多目标强化学习机器人导航方法，实现上下文感知的人类偏好理解。 reinforcement learning large language model

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

#题目一句话要点标签🔗⭐
20 OmniVLN: Omnidirectional 3D Perception and Token-Efficient LLM Reasoning for Visual-Language Navigation across Air and Ground Platforms OmniVLN：面向空地平台视觉语言导航的3D全向感知与高效LLM推理 egocentric multimodal

⬅️ 返回 cs.RO 首页 · 🏠 返回主页