Prof. Xin Xu, Wuhan University of Science and Technology, China

Xin Xu (Senior Member, IEEE) received B.S. and Ph.D. degrees in computer science and engineering from Shanghai Jiao Tong University, China. He is currently a Full Professor at the School of Computer Science and Technology, Wuhan University of Science and Technology, China. His current research interests include artificial intelligence, computer vision, and image processing. More specifically, his research areas focus on building a hierarchical person re-identification architecture including detection and recognition for nighttime surveillance scenarios. He has published more than 160 papers. His publication was shortlisted as the Best Paper Finalist of the IEEE International Conference on Multimedia and Expo (ICME) in 2021 and the Spotlight Paper of the International Conference on Machine Learning (ICML) in 2025. He serves as the Academic Editor for several prestigious journals and delivers Keynote speeches at international conferences.

Speech Title: Fine-grained data selection for re-identification task

Abstract: AI's success is largely attributed to data intelligence and machine learning, which extracts knowledge, patterns, and models from data. One of the problems that big data and model have faced is that, while powerful and easy to demonstrate, it has been a challenge to make it widely available in a variety of scenarios. In many vertical domain, utilizing fully supervised/weakly supervised models usually perform better than using large models directly. To effectively address practical problems in vertical domain and promote AI's integration with the real economy, it is urgent to fully cooperation between big models and data selection.