Xiaobo Xia

Xiaobo Xia

About Me

I am a Postdoctoral Fellow at National University of Singapore, supervised by Prof. Tat-Seng Chua. Prior to this, I obtained my Ph.D. degree at The University of Sydney advised by Prof. Tongliang Liu. Before that, I got my B.E. degree at Xidian University, feeling fortunate to learn from Prof. Nannan Wang. I was a visiting scholar at Yale University. My research area is data-centric AI. The main research interests include trustworthy and efficient AI, AI-generated content, and scientific machine learning. I am a recipient of Google PhD Fellowship 2022.

My general research interests lie in data-centric AI, with a particular emphasis on the following topics:

Trustworthy and Efficient AI:
- Learning with noisy labels (multi-class) [NeurIPS'19, NeurIPS'20, ICLR'21, ICML'21, ICLR'22, CVPR'22, TPAMI'23, ICLR'23, ICCV'23, TPAMI'24]
- Learning with noisy labels (multi-label) [NeurIPS'22, ICCV'23]
- Out-of-distribution detection/generalization [NeurIPS'22, ICLR'23, CVPR'23, NeurIPS'23]
- Learning from crowds [TPAMI’24]
- Data-efficient kernel mean estimation [KDD'22]
- Coreset selection [ICLR'23]
AI-Generated Content:
- Human motion generation [NeurIPS'22, ICCV'23]
- Large language models [ICLR'24, Preprint'23b]
Scientific Machine Learning:
- AI for physics