cs.RO（2025-12-20）

📊 共 8 篇论文

🎯 兴趣领域导航

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

🔬 支柱一：机器人控制 (Robot Control) (5 篇)

#题目一句话要点标签🔗⭐
1 STORM: Search-Guided Generative World Models for Robotic Manipulation STORM：面向机器人操作的搜索引导生成世界模型 manipulation world model latent dynamics
2 AOMGen: Photoreal, Physics-Consistent Demonstration Generation for Articulated Object Manipulation AOMGen：用于铰接物体操作的逼真、物理一致的演示数据生成框架 manipulation world model vision-language-action
3 Learning Semantic Atomic Skills for Multi-Task Robotic Manipulation AtomSkill：面向多任务机器人操作，学习语义原子技能以提升泛化性 manipulation imitation learning contrastive learning
4 On Swarm Leader Identification using Probing Policies 提出基于交互式探测的群体机器人领导者识别方法，解决对抗环境下的领导者隐藏问题。 sim-to-real reinforcement learning deep reinforcement learning
5 Deterministic Reconstruction of Tennis Serve Mechanics: From Aerodynamic Constraints to Internal Torques via Rigid-Body Dynamics 提出基于刚体动力学的网球发球力学确定性重建方法，从空气动力学约束推导内部扭矩。 trajectory optimization

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

#题目一句话要点标签🔗⭐
6 Reinforcement Learning Position Control of a Quadrotor Using Soft Actor-Critic (SAC) 提出基于软演员-评论家算法的四旋翼推力矢量强化学习位置控制方法 reinforcement learning SAC
7 Dynamic Entropy Tuning in Reinforcement Learning Low-Level Quadcopter Control: Stochasticity vs Determinism 提出基于动态熵调整的强化学习四旋翼控制方法，提升探索效率并避免灾难性遗忘。 reinforcement learning SAC TD3

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

#题目一句话要点标签🔗⭐
8 LLaViDA: A Large Language Vision Driving Assistant for Explicit Reasoning and Enhanced Trajectory Planning 提出LLaViDA，利用视觉语言模型增强自动驾驶轨迹规划的推理能力。 scene understanding motion prediction chain-of-thought

⬅️ 返回 cs.RO 首页 · 🏠 返回主页