cs.RO（2026-03-16）

📊 共 27 篇论文 | 🔗 8 篇有代码

🎯 兴趣领域导航

支柱一：机器人控制 (Robot Control) (19 🔗7) 支柱三：空间感知与语义 (Perception & Semantics) (3) 支柱二：RL算法与架构 (RL & Architecture) (3 🔗1) 支柱九：具身大模型 (Embodied Foundation Models) (2)

🔬 支柱一：机器人控制 (Robot Control) (19 篇)

#题目一句话要点标签🔗⭐
1 Simulation Distillation: Pretraining World Models in Simulation for Rapid Real-World Adaptation SimDist：通过模拟器预训练世界模型，实现快速的真实世界适应 quadruped locomotion manipulation
2 AnoleVLA: Lightweight Vision-Language-Action Model with Deep State Space Models for Mobile Manipulation 提出AnoleVLA，一种基于深度状态空间模型，用于移动操作的轻量级视觉-语言-动作模型。 manipulation mobile manipulation state space model
3 MoE-ACT: Scaling Multi-Task Bimanual Manipulation with Sparse Language-Conditioned Mixture-of-Experts Transformers 提出MoE-ACT，通过稀疏MoE Transformer提升多任务双臂操作模仿学习性能。 manipulation bi-manual dual-arm ✅
4 HALO:Closing Sim-to-Real Gap for Heavy-loaded Humanoid Agile Motion Skills via Differentiable Simulation 提出HALO框架，通过可微仿真解决重载人形机器人敏捷运动技能的Sim-to-Real问题 humanoid humanoid robot sim-to-real ✅
5 NavThinker: Action-Conditioned World Models for Coupled Prediction and Planning in Social Navigation NavThinker：基于动作条件世界模型的社交导航耦合预测与规划 Unitree reinforcement learning PPO ✅
6 HapticVLA: Contact-Rich Manipulation via Vision-Language-Action Model without Inference-Time Tactile Sensing 提出HapticVLA，无需推理时触觉传感实现富接触操作 manipulation flow matching distillation
7 CycleRL: Sim-to-Real Deep Reinforcement Learning for Robust Autonomous Bicycle Control CycleRL：用于稳健自主自行车控制的Sim-to-Real深度强化学习框架 sim-to-real domain randomization reinforcement learning ✅
8 From Passive Observer to Active Critic: Reinforcement Learning Elicits Process Reasoning for Robotic Manipulation PRIMO R1：强化学习驱动视频MLLM进行机器人操作过程推理与监督 humanoid manipulation reinforcement learning
9 Ego to World: Collaborative Spatial Reasoning in Embodied Systems via Reinforcement Learning 提出CoRL框架，解决具身多智能体系统中基于强化学习的协同空间推理问题 manipulation reinforcement learning scene understanding
10 KiRAS: Keyframe Guided Self-Imitation for Robust and Adaptive Skill Learning in Quadruped Robots KiRAS：基于关键帧引导的自模仿学习，提升四足机器人复杂地形技能泛化性 quadruped locomotion Unitree
11 ForceVLA2: Unleashing Hybrid Force-Position Control with Force Awareness for Contact-Rich Manipulation 提出ForceVLA2以解决接触丰富操作中的力感知问题 manipulation vision-language-action VLA
12 Exploring the dynamic properties and motion reproducibility of a small upper-body humanoid robot with 13-DOF pneumatic actuation for data-driven control 针对气动人形机器人，提出基于数据驱动的控制方法，提升轨迹跟踪精度。 humanoid humanoid robot
13 RoCo Challenge at AAAI 2026: Benchmarking Robotic Collaborative Manipulation for Assembly Towards Industrial Automation RoCo挑战赛：面向工业自动化的机器人协同装配操作基准测试 manipulation dual-arm VLA ✅
14 Master Micro Residual Correction with Adaptive Tactile Fusion and Force-Mixed Control for Contact-Rich Manipulation 提出M2-ResiPolicy，通过触觉自适应融合和力混合控制，提升接触式操作的微残差校正能力。 manipulation imitation learning diffusion policy
15 End-to-End Dexterous Grasp Learning from Single-View Point Clouds via a Multi-Object Scene Dataset 提出DGS-Net，解决多物体场景下单目点云的灵巧抓取学习问题 manipulation grasp prediction penetration ✅
16 ReMAP-DP: Reprojected Multi-view Aligned PointMaps for Diffusion Policy ReMAP-DP：利用重投影多视角对齐点云图的扩散策略，提升机器人操作精度 manipulation diffusion policy ✅
17 Confusion-Aware In-Context-Learning for Vision-Language Models in Robotic Manipulation 提出Confusion-Aware In-Context Learning，提升VLM在机器人操作中对易混淆物体的识别能力。 manipulation
18 A Unified Calibration Framework for Coordinate and Kinematic Parameters in Dual-Arm Robots 提出双臂机器人坐标与运动学参数统一标定框架，提升协作精度 dual-arm
19 AeroGrab: A Unified Framework for Aerial Grasping in Cluttered Environments AeroGrab：提出统一框架，解决复杂环境下空中抓取的可靠性问题 manipulation

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

#题目一句话要点标签🔗⭐
20 NavGSim: High-Fidelity Gaussian Splatting Simulator for Large-Scale Navigation NavGSim：用于大规模导航的高保真高斯溅射模拟器 3D gaussian splatting gaussian splatting splatting
21 LiDAR-EVS: Enhance Extrapolated View Synthesis for 3D Gaussian Splatting with Pseudo-LiDAR Supervision LiDAR-EVS：利用伪激光雷达监督增强3D高斯溅射外推视图合成，用于自动驾驶 3D gaussian splatting 3DGS gaussian splatting
22 Coupled Particle Filters for Robust Affordance Estimation 提出耦合粒子滤波方法，用于解决机器人操作中的稳健可供性估计问题 affordance

🔬 支柱二：RL算法与架构 (RL & Architecture) (3 篇)

#题目一句话要点标签🔗⭐
23 MA-VLCM: A Vision Language Critic Model for Value Estimation of Policies in Multi-Agent Team Settings 提出MA-VLCM，利用视觉语言模型提升多智能体强化学习策略价值估计的样本效率。 reinforcement learning vision-language-action VLA
24 PerlAD: Towards Enhanced Closed-loop End-to-end Autonomous Driving with Pseudo-simulation-based Reinforcement Learning PerlAD：基于伪模拟强化学习的端到端闭环自动驾驶 reinforcement learning imitation learning world model
25 ViSA: Visited-State Augmentation for Generalized Goal-Space Contrastive Reinforcement Learning 提出ViSA，通过访问状态增强提升目标空间对比强化学习的泛化能力 reinforcement learning contrastive learning ✅

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

#题目一句话要点标签🔗⭐
26 Learning from Mistakes: Post-Training for Driving VLA with Takeover Data TakeVLA：通过接管数据后训练，提升端到端自动驾驶VLA模型的安全性和性能。 vision-language-action VLA
27 GraspALL: Adaptive Structural Compensation from Illumination Variation for Robotic Garment Grasping in Any Low-Light Conditions GraspALL：通过光照自适应结构补偿实现任意低光照条件下机器人服装抓取 multimodal

⬅️ 返回 cs.RO 首页 · 🏠 返回主页