Xu Luo

I am a Ph.D. candidate at UESTC, advised by Prof. Jingkuan Song. My research aims to bridge the gap between virtual AI and the physical world by developing general-purpose robots, focusing on the creation of generalist policies that empower them to robustly perceive, dynamically interact, and continuously adapt—ultimately enabling them to tackle any task, in any environment.

Email  /  CV  /  Google Scholar  /  Github

Identifying shortcut learning as a key impediment to the generalization of generalist robot policies and providing a comprehensive analysis.

Empirically proving the disentanglement of training and adaptation algorithms in few-shot classification, and performing interesting analysis of each phase that leads to the discovery of several important observations.

Revealing and analyzing the channel bias problem that we found critical in few-shot learning, through a simple channel-wise feature transformation applied only at test time.

Identifying image background as a shortcut knowledge ungeneralizable beyond training categories in Few-Shot Learning. A novel framework, COSOC, is designed to tackle this problem.

Conference reviewer

This well-designed template is borrowed from this guy