← Search

Xuanyu Xiang

2 accepted papers

2026

HPS: Hyperspherical Parameter Sharing for Efficient Multi-Agent Reinforcement Learning

ICML 2026poster

Parameter Sharing (PS) is widely used to improve efficiency in Multi-Agent Reinforcement Learning (MARL), but it can limit behavioral diversity and degrade performance. This limitation stems from gradient conflicts among agents on shared weights, which hinders effective policy learning. To fully cha…

Cited by 0SourceScholar
2024

MonoCD: Monocular 3D Object Detection with Complementary Depths

CVPR 2024poster

Monocular 3D object detection has attracted widespread attention due to its potential to accurately obtain object 3D localization from a single image at a low cost. Depth estimation is an essential but challenging subtask of monocular 3D object detection due to the ill-posedness of 2D to 3D mapping.…