cs.RO（2026-08-04）

📊 共 25 篇论文 | 🔗 2 篇有代码

🎯 兴趣领域导航

支柱一：机器人控制 (Robot Control) (19 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (2) 支柱八：物理动画 (Physics-based Animation) (2) 支柱九：具身大模型 (Embodied Foundation Models) (1) 支柱六：视频提取与匹配 (Video Extraction) (1)

🔬 支柱一：机器人控制 (Robot Control) (19 篇)

#题目一句话要点标签🔗⭐
1 RoboReact: Agentic Skill Distillation from Generated Egocentric Videos for Generalizable Whole-Body Manipulation 提出RoboReact以解决人形机器人技能获取成本高的问题 humanoid humanoid robot whole-body control
2 PFM-HR: Pose Flow Matching for Humanoid Robots 提出PFM-HR以解决人形机器人运动跟踪中的时间先验问题 humanoid humanoid robot motion tracking
3 Bimanual Manipulation Within an 8 GB Budget: Zero-Copy Sensing and Quantized ACT on an Entry-Level Jetson 提出基于零拷贝感知的双手操作策略以适应嵌入式硬件 manipulation bi-manual bimanual manipulation
4 Structure-Aware Robust Fine-Tuning: Defending Vision-Language-Action Robots Against Physical Attention Hijacking 提出结构感知鲁棒微调以防御视觉-语言-动作机器人对抗攻击 manipulation geometric consistency vision-language-action
5 Track4Action: Distilling World-Centric 3D Tracker into Vision-Language-Action Policies 提出Track4Action以解决3D跟踪与视觉语言行动政策的对齐问题 bi-manual flow matching vision-language-action ✅
6 Learning Context-Aware Motion Priors for Humanoid Control 提出上下文感知运动先验以解决人形机器人控制问题 humanoid humanoid control policy learning
7 Unified Visuomotor Targets: Supervising VLAs Beyond Physical Actions 提出统一视觉运动目标以解决VLA模型训练效率问题 manipulation bi-manual bimanual manipulation
8 Shooting for Contact: Contact-Implicit Multiple Shooting for Dynamic Motion Retargeting 提出接触隐式多重射击框架以解决动态运动重定向问题 sim-to-real Unitree reinforcement learning
9 Continue or Replan? Bernoulli-Continuation Policy Learning for Adaptive Horizon Execution 提出Bernoulli-Continuation策略以解决固定执行视角的局限性 manipulation reinforcement learning policy learning
10 A Hierarchical Approach to Imitation Learning for Manipulation Tasks Requiring Time Varying Forces 提出DPA-FTG以解决动态交互中的控制延迟问题 manipulation bi-manual imitation learning
11 CUDA MPC: A GPU-Native Solver for Model Predictive Control 提出CUDA MPC以解决快速动态系统中的模型预测控制问题 MPC model predictive control
12 EmbodiedVAE: Disentangled Video VAE for Efficient and Controllable Embodied Manipulation 提出EmbodiedVAE以解决现有LDM在机器人操作中的控制问题 manipulation world model world models
13 GORDON: Graph-based Object-centric Rewards for Decomposition of Long-Horizon Manipulation 提出GORDON框架以解决长时间操作中的奖励设计问题 manipulation reinforcement learning policy learning
14 EvoHIL: Self-Evolving Reward and Flow-Matched Policy Optimization for Robust Human-in-the-Loop Reinforcement Learning 提出EvoHIL以解决人机协同强化学习中的动态适应问题 manipulation reinforcement learning flow matching ✅
15 Stochastic Multiple Shooting Trajectory Optimization via Sequential Local Policy Evaluation 提出随机多重发射轨迹优化方法以提高样本效率 trajectory optimization reinforcement learning
16 Accelerating Human-Aware Robot Trajectory Generation via Diffusion and Consistency Distillation 提出基于扩散与一致性蒸馏的机器人轨迹生成方法以解决人机交互中的安全问题 motion planning distillation
17 LiLa-WAM: Lightweight Latent Reasoning World-Action Model for Robotic Manipulation 提出LiLa-WAM以解决机器人控制中的计算开销问题 manipulation
18 A Wearable Stiffness-Rendering Haptic Device with a Honeycomb Jamming Mechanism for Bilateral Teleoperation 提出可穿戴的蜂窝夹紧机制触觉设备以解决双向遥操作中的运动反馈问题 teleoperation
19 GraspMeanFlow: SE(3)-Equivariant MeanFlow for Few-Step 6-DoF Grasp Generation 提出GraspMeanFlow以解决实时6自由度抓取生成问题 manipulation flow matching

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

#题目一句话要点标签🔗⭐
20 How Should Vision-Language-Action Models Use Proprioceptive State? 提出有效的状态输入方式以优化视觉-语言-动作模型 flow matching vision-language-action VLA
21 Lightweight 3D Object Detection via Mamba-Based Knowledge Distillation 提出基于Mamba的知识蒸馏框架以实现轻量级3D目标检测 Mamba distillation

🔬 支柱八：物理动画 (Physics-based Animation) (2 篇)

#题目一句话要点标签🔗⭐
22 Tired Actor: Fatigue-Informed Character Control 提出疲劳模型以提升角色动画的自然性与鲁棒性 physics-based character character animation character control
23 PLS-Calib: A Partial Least Squares Framework for Event Camera and Odometry Calibration under Ground Motion Constraints 提出PLS-Calib以解决地面约束机器人旋转标定问题 spatiotemporal

🔬 支柱九：具身大模型 (Embodied Foundation Models) (1 篇)

#题目一句话要点标签🔗⭐
24 PACE: Adaptive Budget Allocation for Time-Efficient Embodied Planning 提出PACE框架以解决时间效率低下的规划问题 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🔗⭐
25 DigitCode: Symbolic Tokenization of Hand Motion by Anatomical Units 提出DigitCode以解决手部运动符号化表示问题 MANO

⬅️ 返回 cs.RO 首页 · 🏠 返回主页