cs.RO（2025-09-19）

📊 共 24 篇论文 | 🔗 5 篇有代码

🎯 兴趣领域导航

支柱一：机器人控制 (Robot Control) (17 🔗4) 支柱三：空间感知与语义 (Perception & Semantics) (3) 支柱二：RL算法与架构 (RL & Architecture) (2 🔗1) 支柱九：具身大模型 (Embodied Foundation Models) (2)

🔬 支柱一：机器人控制 (Robot Control) (17 篇)

#题目一句话要点标签🔗⭐
1 PRIMT: Preference-based Reinforcement Learning with Multimodal Feedback and Trajectory Synthesis from Foundation Models 提出PRIMT以解决偏好强化学习中的人类输入依赖与奖励学习难题 locomotion manipulation reinforcement learning
2 Latent Conditioned Loco-Manipulation Using Motion Priors 提出基于运动先验的潜在条件Loco-Manipulation方法，用于人形和四足机器人的复杂操作任务 quadruped humanoid manipulation ✅
3 A Vision-Language-Action-Critic Model for Robotic Real-World Reinforcement Learning 提出VLAC模型，解决机器人真实世界强化学习中奖励稀疏和探索低效问题。 manipulation reinforcement learning vision-language-action
4 ORB: Operating Room Bot, Automating Operating Room Logistics through Mobile Manipulation ORB：提出一种基于移动操作的机器人系统，用于自动化手术室物流。 manipulation mobile manipulation trajectory optimization
5 Dynamic Objects Relocalization in Changing Environments with Flow Matching 提出FlowMaps，利用Flow Matching预测动态环境中物体位置，解决未知重定位问题 motion planning flow matching human-object interaction ✅
6 Momentum-constrained Hybrid Heuristic Trajectory Optimization Framework with Residual-enhanced DRL for Visually Impaired Scenarios 针对视障人士，提出动量约束混合启发式轨迹优化框架，结合残差增强DRL。 trajectory optimization reinforcement learning deep reinforcement learning
7 A Framework for Optimal Ankle Design of Humanoid Robots 提出人形机器人踝关节优化设计框架，提升地面交互性能。 humanoid humanoid robot
8 DSPv2: Improved Dense Policy for Effective and Generalizable Whole-body Mobile Manipulation DSPv2：改进的密集策略，用于有效且泛化的全身移动操作 manipulation mobile manipulation ✅
9 Improving Robotic Manipulation with Efficient Geometry-Aware Vision Encoder 提出高效几何感知视觉编码器eVGGT，提升机器人操作性能 manipulation bi-manual imitation learning
10 Reward Evolution with Graph-of-Thoughts: A Bi-Level Language Model Framework for Reinforcement Learning 提出RE-GoT框架，利用图推理和视觉反馈实现强化学习中奖励函数的自动进化。 manipulation reinforcement learning reward design
11 Right-Side-Out: Learning Zero-Shot Sim-to-Real Garment Reversal 提出Right-Side-Out框架，解决服装翻转中零样本Sim-to-Real难题 manipulation bi-manual sim-to-real
12 An MPC framework for efficient navigation of mobile robots in cluttered environments 提出一种MPC框架，用于移动机器人在复杂环境中高效导航 MPC model predictive control trajectory optimization
13 GP3: A 3D Geometry-Aware Policy with Multi-View Images for Robotic Manipulation GP3：利用多视角图像和3D几何感知的机器人操作策略 manipulation
14 Imagination at Inference: Synthesizing In-Hand Views for Robust Visuomotor Policy Inference 提出想象感知以解决机器人操作中的视角限制问题 manipulation Unitree egocentric
15 I-FailSense: Towards General Robotic Failure Detection with Vision-Language Models I-FailSense：利用视觉-语言模型实现通用机器人故障检测 manipulation language conditioned ✅
16 A Matter of Height: The Impact of a Robotic Object on Human Compliance 研究机器人高度对人类顺从性的影响，发现与人际互动相反的模式 humanoid humanoid robot
17 Compose by Focus: Scene Graph-based Atomic Skills 提出基于场景图的原子技能学习框架，提升机器人组合泛化能力 manipulation imitation learning

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

#题目一句话要点标签🔗⭐
18 Agentic Aerial Cinematography: From Dialogue Cues to Cinematic Trajectories ACDC：提出一种基于对话提示的自主无人机电影摄影系统 open-vocabulary open vocabulary embodied AI
19 FiLM-Nav: Efficient and Generalizable Navigation via VLM Fine-tuning FiLM-Nav：通过VLM微调实现高效且可泛化的导航 open-vocabulary open vocabulary foundation model
20 Omni-LIVO: Robust RGB-Colored Multi-Camera Visual-Inertial-LiDAR Odometry via Photometric Migration and ESIKF Fusion Omni-LIVO：基于光度迁移和ESIKF融合的鲁棒RGB彩色多相机视觉-惯性-激光雷达里程计 visual odometry LIO

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

#题目一句话要点标签🔗⭐
21 CoReVLA: A Dual-Stage End-to-End Autonomous Driving Framework for Long-Tail Scenarios via Collect-and-Refine CoReVLA：通过收集与优化双阶段学习，提升长尾场景下自动驾驶性能 DPO direct preference optimization VLA ✅
22 End-to-end RL Improves Dexterous Grasping Policies 提出解耦模拟器与强化学习的策略，提升灵巧抓取的端到端强化学习性能 PPO distillation

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

#题目一句话要点标签🔗⭐
23 Defining and Monitoring Complex Robot Activities via LLMs and Symbolic Reasoning 利用LLM与符号推理定义和监控复杂机器人活动 large language model
24 Distribution Estimation for Global Data Association via Approximate Bayesian Inference 提出基于近似贝叶斯推断的全局数据关联方法，解决多模态分布估计问题 multimodal

⬅️ 返回 cs.RO 首页 · 🏠 返回主页