cs.RO（2026-02-12）

📊 共 23 篇论文 | 🔗 6 篇有代码

🎯 兴趣领域导航

支柱一：机器人控制 (Robot Control) (16 🔗5) 支柱二：RL算法与架构 (RL & Architecture) (5 🔗1) 支柱三：空间感知与语义 (Perception & Semantics) (2)

🔬 支柱一：机器人控制 (Robot Control) (16 篇)

#题目一句话要点标签🔗⭐
1 3DGSNav: Enhancing Vision-Language Model Reasoning for Object Navigation via Active 3D Gaussian Splatting 提出3DGSNav，利用主动3D高斯溅射增强视觉语言模型在物体导航中的推理能力。 quadruped 3D gaussian splatting 3DGS ✅
2 ReaDy-Go: Real-to-Sim Dynamic 3D Gaussian Splatting Simulation for Environment-Specific Visual Navigation with Moving Obstacles ReaDy-Go：用于动态环境视觉导航的真实到仿真3D高斯溅射模拟 sim-to-real policy learning 3D gaussian splatting ✅
3 HAIC: Humanoid Agile Object Interaction Control via Dynamics-Aware World Model 提出HAIC框架以解决人形机器人与动态物体交互问题 humanoid humanoid robot world model
4 VLAW: Iterative Co-Improvement of Vision-Language-Action Policy and World Model VLAW：迭代共提升视觉-语言-动作策略与世界模型，提升机器人操作性能 manipulation world model vision-language-action
5 General Humanoid Whole-Body Control via Pretraining and Fast Adaptation FAST：基于预训练和快速自适应的通用人形机器人全身控制框架 humanoid humanoid robot whole-body control
6 When would Vision-Proprioception Policies Fail in Robotic Manipulation? 提出GAP算法，解决机器人操作中视觉-本体感觉策略在运动过渡阶段的失效问题 manipulation dual-arm policy learning
7 HoloBrain-0 Technical Report HoloBrain-0：提出融合机器人先验知识的VLA框架，实现高效真实世界机器人部署 manipulation vision-language-action VLA
8 Accelerating Robotic Reinforcement Learning with Agent Guidance 提出AGPS框架，用多模态Agent指导机器人强化学习，提升样本效率。 manipulation reinforcement learning world model ✅
9 Multi Graph Search for High-Dimensional Robot Motion Planning 提出多图搜索(MGS)算法，解决高维机器人运动规划中效率与一致性难题。 manipulation mobile manipulation motion planning ✅
10 Robot-DIFT: Distilling Diffusion Features for Geometrically Consistent Visuomotor Control Robot-DIFT：通过蒸馏扩散特征实现几何一致的视觉运动控制 manipulation distillation geometric consistency
11 Safety Beyond the Training Data: Robust Out-of-Distribution MPC via Conformalized System Level Synthesis 提出基于保形预测和系统级综合的鲁棒分布外MPC方法，提升模型预测控制的安全性。 MPC model predictive control
12 ViTaS: Visual Tactile Soft Fusion Contrastive Learning for Visuomotor Learning ViTaS：基于视觉触觉软融合对比学习的机器人灵巧操作 manipulation contrastive learning ✅
13 Sub--Riemannian boundary value problems for Optimal Geometric Locomotion 提出基于次黎曼几何的最优几何运动模型，用于解决细长运动体的运动规划问题 locomotion
14 Pack it in: Packing into Partially Filled Containers Through Contact 提出一种接触感知的装箱方法，用于在部分填充容器中放置物品 model predictive control contact-aware
15 Learning to Manipulate Anything: Revealing Data Scaling Laws in Bounding-Box Guided Policies 提出基于边界框引导的策略，揭示语义操作中的数据缩放规律 manipulation diffusion policy
16 Adaptive-Horizon Conflict-Based Search for Closed-Loop Multi-Agent Path Finding 提出自适应时域冲突搜索算法ACCBS，解决闭环多智能体路径规划问题 MPC

🔬 支柱二：RL算法与架构 (RL & Architecture) (5 篇)

#题目一句话要点标签🔗⭐
17 ABot-N0: Technical Report on the VLA Foundation Model for Versatile Embodied Navigation ABot-N0：用于多功能具身导航的VLA基础模型，实现跨任务统一。 flow matching vision-language-action VLA
18 Scaling Verification Can Be More Effective than Scaling Policy Learning for Vision-Language-Action Alignment 提出CoVer：通过验证而非策略学习提升视觉-语言-动作对齐效果 policy learning vision-language-action VLA
19 LDA-1B: Scaling Latent Dynamics Action Model via Universal Embodied Data Ingestion LDA-1B：通过通用具身数据摄取扩展潜在动力学行为模型 behavior cloning world model latent dynamics
20 Any House Any Task: Scalable Long-Horizon Planning for Abstract Human Tasks AHAT：可扩展的家庭机器人长程任务规划框架 reinforcement learning large language model language conditioned
21 AC-MASAC: An Attentive Curriculum Learning Framework for Heterogeneous UAV Swarm Coordination 提出AC-MASAC框架，解决异构无人机集群协同中的非对称依赖和训练难题。 reinforcement learning curriculum learning ✅

🔬 支柱三：空间感知与语义 (Perception & Semantics) (2 篇)

#题目一句话要点标签🔗⭐
22 Affordance-Graphed Task Worlds: Self-Evolving Task Generation for Scalable Embodied Learning 提出基于可供性图的任务世界，用于可扩展具身学习的自进化任务生成。 affordance
23 LAMP: Implicit Language Map for Robot Navigation LAMP：用于机器人导航的隐式语言地图，实现高效精细的路径规划 implicit representation

⬅️ 返回 cs.RO 首页 · 🏠 返回主页