cs.RO（2025-07-07）

📊 共 18 篇论文 | 🔗 3 篇有代码

🎯 兴趣领域导航

支柱一：机器人控制 (Robot Control) (11 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (4) 支柱九：具身大模型 (Embodied Foundation Models) (2) 支柱三：空间感知与语义 (Perception & Semantics) (1)

🔬 支柱一：机器人控制 (Robot Control) (11 篇)

#	题目	一句话要点	标签	🔗	⭐
1	EmbodieDreamer: Advancing Real2Sim2Real Transfer for Policy Training via Embodied World Modeling	EmbodieDreamer：通过具身世界建模提升策略训练的Real2Sim2Real迁移效果	sim2real real2sim reinforcement learning
2	Safe Bimanual Teleoperation with Language-Guided Collision Avoidance	提出一种基于语音引导碰撞避免的安全双臂遥操作系统	whole-body control manipulation bi-manual
3	A Careful Examination of Large Behavior Models for Multitask Dexterous Manipulation	提出基于扩散策略的大型行为模型（LBM），用于多任务灵巧操作，提升泛化性和数据效率。	manipulation dexterous manipulation imitation learning	✅
4	LERa: Replanning with Visual Feedback in Instruction Following	LERa：基于视觉反馈的指令跟随任务重规划方法	manipulation large language model instruction following
5	Monte Carlo Tree Search with Tensor Factorization for Robot Optimization	提出基于张量分解的蒙特卡洛树搜索(TTTS)算法，用于解决机器人优化问题。	legged robot manipulation whole-body manipulation
6	PRISM: Pointcloud Reintegrated Inference via Segmentation and Cross-attention for Manipulation	PRISM：通过分割和交叉注意力进行点云重构推理，用于机器人操作	manipulation imitation learning	✅
7	Training-free Generation of Temporally Consistent Rewards from VLMs	提出T²-VLM，无需训练即可从VLM生成时序一致的机器人操作奖励。	manipulation reinforcement learning embodied AI	✅
8	Interaction-Merged Motion Planning: Effectively Leveraging Diverse Motion Datasets for Robust Planning	提出交互合并运动规划以解决运动数据集利用不足问题	motion planning
9	DRAE: Dynamic Retrieval-Augmented Expert Networks for Lifelong Learning and Task Adaptation in Robotics	提出动态检索增强专家网络DRAE，解决机器人终身学习和任务自适应问题	manipulation reinforcement learning
10	Gaussian Process-Based Active Exploration Strategies in Vision and Touch	提出基于高斯过程距离场的视觉触觉融合主动探索方法，用于机器人感知物体属性。	manipulation
11	Dynamics and multi-stability of a rotor-actuated Twistcar robot with passive steering joint	研究被动转向Twistcar机器人，揭示转子驱动下丰富的非线性动力学与多稳态特性	locomotion

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

#	题目	一句话要点	标签	🔗	⭐
12	IDAGC: Adaptive Generalized Human-Robot Collaboration via Human Intent Estimation and Multimodal Policy Learning	提出IDAGC框架，通过人类意图估计和多模态策略学习实现自适应人机协作。	policy learning predictive model multimodal
13	Action Space Reduction Strategies for Reinforcement Learning in Autonomous Driving	提出动态掩码与相对动作空间缩减策略，提升自动驾驶强化学习效率	reinforcement learning policy learning multimodal
14	NavigScene: Bridging Local Perception and Global Navigation for Beyond-Visual-Range Autonomous Driving	NavigScene：弥合局部感知与全局导航，实现超视距自动驾驶	reinforcement learning direct preference optimization vision-language-action
15	CRED: Counterfactual Reasoning and Environment Design for Active Preference Learning	提出CRED，通过环境设计和反事实推理提升主动偏好学习效果	preference learning

🔬 支柱九：具身大模型 (Embodied Foundation Models) (2 篇)

#	题目	一句话要点	标签	🔗	⭐
16	StreamVLN: Streaming Vision-and-Language Navigation via SlowFast Context Modeling	提出StreamVLN，通过慢速-快速上下文建模实现低延迟的流式视觉语言导航。	VLN large language model
17	VerifyLLM: LLM-Based Pre-Execution Task Plan Verification for Robots	VerifyLLM：基于LLM的机器人任务规划预执行验证	large language model

🔬 支柱三：空间感知与语义 (Perception & Semantics) (1 篇)

#	题目	一句话要点	标签	🔗	⭐
18	MOSU: Autonomous Long-range Robot Navigation with Multi-modal Scene Understanding	MOSU：基于多模态场景理解的自主长距离机器人导航系统	scene understanding traversability multimodal

⬅️ 返回 cs.RO 首页 · 🏠 返回主页