cs.RO（2026-03-03）

📊 共 25 篇论文 | 🔗 1 篇有代码

🎯 兴趣领域导航

支柱一：机器人控制 (Robot Control) (19 🔗1) 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) 支柱二：RL算法与架构 (RL & Architecture) (1)

🔬 支柱一：机器人控制 (Robot Control) (19 篇)

#	题目	一句话要点	标签	🔗	⭐
1	ULTRA: Unified Multimodal Control for Autonomous Humanoid Whole-Body Loco-Manipulation	提出ULTRA框架，实现基于多模态感知的自主人形机器人全身运动控制。	humanoid manipulation loco-manipulation
2	HoMMI: Learning Whole-Body Mobile Manipulation from Human Demonstrations	HoMMI：从人类演示中学习全身移动操作	whole-body control manipulation mobile manipulation
3	Rhythm: Learning Interactive Whole-Body Control for Dual Humanoids	Rhythm：提出交互式全身控制框架，实现双人形机器人协同	humanoid whole-body control Unitree
4	CMoE: Contrastive Mixture of Experts for Motion Control and Terrain Adaptation of Humanoid Robots	提出CMoE，通过对比学习提升人形机器人运动控制和地形适应能力	humanoid humanoid robot Unitree
5	Watch Your Step: Learning Semantically-Guided Locomotion in Cluttered Environment	SemLoco：提出一种语义引导的强化学习框架，用于复杂环境中四足机器人的安全运动。	legged robot locomotion reinforcement learning
6	ACE-Brain-0: Spatial Intelligence as a Shared Scaffold for Universal Embodiments	ACE-Brain-0：提出基于空间智能的通用具身智能框架，解决跨异构载体的泛化难题。	manipulation cross-embodiment embodiment transfer
7	DreamFlow: Local Navigation Beyond Observation via Conditional Flow Matching in the Latent Space	DreamFlow：基于条件流匹配的局部导航，扩展机器人感知范围	quadruped reinforcement learning deep reinforcement learning
8	Learning Object-Centric Spatial Reasoning for Sequential Manipulation in Cluttered Environments	Unveiler：面向杂乱环境顺序操作，学习以物体为中心的空间推理	manipulation PPO imitation learning
9	Self-supervised Domain Adaptation for Visual 3D Pose Estimation of Nano-drone Racing Gates by Enforcing Geometric Consistency	提出自监督领域适应方法以解决纳米无人机赛道门的3D姿态估计问题	sim-to-real geometric consistency
10	IMR-LLM: Industrial Multi-Robot Task Planning and Program Generation using Large Language Models	IMR-LLM：基于大语言模型的工业多机器人任务规划与程序生成框架	manipulation large language model
11	Look Forward to Walk Backward: Efficient Terrain Memory for Backward Locomotion with Forward Vision	提出LF2WB框架，利用前视记忆实现腿足机器人无后视的后向敏捷运动	legged robot locomotion egocentric
12	Uni-Skill: Building Self-Evolving Skill Repository for Generalizable Robotic Manipulation	Uni-Skill：构建自进化技能库，提升机器人操作的泛化性	manipulation foundation model
13	How to Peel with a Knife: Aligning Fine-Grained Manipulation with Human Preference	提出一种基于模仿学习和偏好学习的机器人削皮方法，提升精细操作性能。	manipulation imitation learning
14	cuNRTO: GPU-Accelerated Nonlinear Robust Trajectory Optimization	提出cuNRTO框架，加速GPU上的非线性鲁棒轨迹优化	trajectory optimization
15	Give me scissors: Collision-Free Dual-Arm Surgical Assistive Robot for Instrument Delivery	提出基于视觉-语言模型的双臂手术辅助机器人，实现无碰撞器械递送	dual-arm	✅
16	Tracing Back Error Sources to Explain and Mitigate Pose Estimation Failures	提出基于误差溯源的模块化框架，提升机器人抓取中位姿估计的鲁棒性	manipulation foundation model
17	CASSR: Continuous A-Star Search through Reachability for real time footstep planning	CASSR：基于可达性分析的连续A*搜索，实现机器人实时步态规划	biped locomotion
18	RL-Based Coverage Path Planning for Deformable Objects on 3D Surfaces	提出基于强化学习的覆盖路径规划方法，用于3D表面上可变形物体的操作。	manipulation reinforcement learning
19	Design, Modeling and Direction Control of a Wire-Driven Robotic Fish Based on a 2-DoF Crank-Slider Mechanism	提出基于2自由度曲柄滑块机构的线驱动仿生鱼，实现高速游泳和灵活转向	locomotion

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

#	题目	一句话要点	标签	🔗	⭐
20	MA-CoNav: A Master-Slave Multi-Agent Framework with Hierarchical Collaboration and Dual-Level Reflection for Long-Horizon Embodied VLN	提出MA-CoNav多智能体框架，解决长程具身VLN任务中的感知失真和决策漂移问题。	VLN
21	From Language to Action: Can LLM-Based Agents Be Used for Embodied Robot Cognition?	提出基于LLM的认知机器人架构，用于家庭环境中的具身智能任务	large language model instruction following
22	Agentic Self-Evolutionary Replanning for Embodied Navigation	提出SERP，通过智能体自进化重规划提升具身导航的鲁棒性与效率。	large language model 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	🔗	⭐
23	CoFL: Continuous Flow Fields for Language-Conditioned Navigation	CoFL：提出基于连续流场的语言条件导航方法，实现高效端到端控制。	semantic map language conditioned
24	PathSpace: Rapid continuous map approximation for efficient SLAM using B-Splines in constrained environments	PathSpace：利用B样条进行快速连续地图近似，实现约束环境下的高效SLAM	visual SLAM

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

#	题目	一句话要点	标签	🔗	⭐
25	Generative adversarial imitation learning for robot swarms: Learning from human demonstrations and trained policies	提出基于生成对抗模仿学习的机器人集群控制方法，从人类演示和策略模型中学习。	PPO imitation learning

⬅️ 返回 cs.RO 首页 · 🏠 返回主页