Prof. Tao Lei
Shaanxi University of Science and Technology, China

Tao Lei is a professor and doctoral supervisor at Shaanxi University of Science and Technology. He is alos the vice dean of the School of Electronic Information and Artificial Intelligence, and Senior Member of IEEE/CCF/CSIG. He is selected from the Shaanxi Provincial High level Talent Program, Shaanxi Provincial Outstanding Youth, Stanford Top 2% Global Scientists List, etc. He is a deputy editor, editorial board member, guest editor, etc. for 7 journals, and serves as conference chairman, technical committee chairman, publicity chairman, reward committee chairman, branch chairman, etc. in more than 20 international conferences. His main research areas are computer vision, machine learning, etc. At present, He has published 4 collections of specialized/authored and conference papers, and have published over 100 papers in international journals and conferences such as IEEE TIP, IEEE TMI, IEEE TFS, IEEE TGRS, and IJCAI. Among them, 9 papers are ESI highly cited papers. His Google Academic Citation has exceeded 4700. He hosted many projects such as the National Natural Science Foundation of China (5 projects), Shaanxi Provincial Outstanding Youth Fund, and Shaanxi Provincial Key Research and Development Program. He won the second prize of Shaanxi Province Science and Technology Award and the first prize of Gansu Province Higher Education Research Excellent Achievement Award as the first complete person.

Speech Title "Semi-supervised Medical Image Segmentation under Limited Labeled Data"

Abstract: Medical image segmentation is a key technology in the field of intelligent image analysis. At present, a large number of research results on medical image segmentation have been reported and used for smart medicine. However, the current mainstream medical image segmentation methods still face the following challenges. Firstly, accurate segmentation of medical images is difficult due to high noise and low contrast. Secondly, mainstream medical image segmentation models have a large number of parameters and slow inference speed, making it difficult to deploy on low resource devices. Finally, pixel-level annotation of medical images is very expensive and requires professional knowledge. To address these problems, our team focuses on semi-supervised medical image segmentation and its applications. We have proposed some novelty semi-supervised network models for medical image segmentation under limited labeled samples, and they show good performance for some complex tasks.