10 papers
FlashNav: Training Deployable Robot Navigation Policies in Seconds
Shanze Wang, Yiwei Qian, Xinming Zhang +8
Training Deep Reinforcement Learning (DRL) navigation policies for different robot configurations remains time-consuming. We present FlashNav, a GPU-based framework that trains rob…
Do We Really Need Immediate Resets? Rethinking Collision Handling for Efficient Robot Navigation
Shanze Wang, Xinming Zhang, Siwei Cheng +4
Should a single collision necessarily terminate an entire navigation episode? In most deep reinforcement learning (DRL) frameworks for robot navigation, this remains the standard p…
Learning from Demonstration with Failure Awareness for Safe Robot Navigation
Xianghui Wang, Siwei Cheng, Shanze Wang +3
Learning from demonstration is widely used for robot navigation, yet it suffers from a fundamental limitation: demonstrations consist predominantly of successful behaviors and prov…
FSUNav: A Cerebrum-Cerebellum Architecture for Fast, Safe, and Universal Zero-Shot Goal-Oriented Navigation
Mingao Tan, Yiyang Li, Shanze Wang +2
Current vision-language navigation methods face substantial bottlenecks regarding heterogeneous robot compatibility, real-time performance, and navigation safety. Furthermore, they…
CeRLP: A Cross-embodiment Robot Local Planning Framework for Visual Navigation
Haoyu Xi, Mingao Tan, Xinming Zhang +5
Visual navigation for cross-embodiment robots is challenging due to variations in robot and camera configurations, which can lead to the failure of navigation tasks. Previous appro…
FastDSAC: Unlocking the Potential of Maximum Entropy RL in High-Dimensional Humanoid Control
Jun Xue, Junze Wang, Shanze Wang +3
Scaling Maximum Entropy Reinforcement Learning (RL) to high-dimensional humanoid control remains a fundamental challenge, as the ''curse of dimensionality'' induces severe explorat…
The Trace Is the State: Exact Credit Assignment for LLM Agent Teams
Yanjun Chen, Yirong Sun, Hanlin Wang +5
Credit assignment for a team of LLM agents, what each message was worth, has mostly been treated as prediction: a learned critic, a trajectory-level score, or an agent ablation sta…
MA-ROESL: Motion-aware Rapid Reward Optimization for Efficient Robot Skill Learning from Single Videos
Xianghui Wang, Xinming Zhang, Yanjun Chen +2
Vision-language models (VLMs) have demonstrated excellent high-level planning capabilities, enabling locomotion skill learning from video demonstrations without the need for meticu…
Rethinking Soft Actor-Critic in High-Dimensional Action Spaces: The Cost of Ignoring Distribution Shift
Yanjun Chen, Xinming Zhang, Xianghui Wang +3
Soft Actor-Critic algorithm is widely recognized for its robust performance across a range of deep reinforcement learning tasks, where it leverages the tanh transformation to const…
Achieving Stable High-Speed Locomotion for Humanoid Robots with Deep Reinforcement Learning
Xinming Zhang, Xianghui Wang, Lerong Zhang +3
Humanoid robots offer significant versatility for performing a wide range of tasks, yet their basic ability to walk and run, especially at high velocities, remains a challenge. Thi…