Abstract

In this talk, we will discuss several issues related to thermal face analysis. Targeting at developing automatic monitoring or access control systems with the consideration of privacy preserving, we work on images/videos captured in the infrared spectrum. Given images/videos captured in arbitrary poses and distances, we first develop a face detection module dedicated to thermal faces. In order to facilitate more accurate thermal face recognition, for the detected face regions we develop a facial landmark detection module based on U-Net and multi-task learning. With detected facial landmarks, head poses can be estimated, and faces can be transformed into the frontal view. Finally, we develop a thermal face recognition system by transforming thermal faces into that in the visible spectrum, and then recognizing the transformed faces. We will introduce the research topics about thermal faces, and describe details of the aforementioned technical components in this talk.

Bio:

Wei-Ta Chu received the B.S. and M.S. degrees in Computer Science from National Chi Nan University, Taiwan, in 2000 and 2002, and received the Ph.D. degree in Computer Science from National Taiwan University, Taiwan, in 2006. He was a Professor in National Chung Cheng University from 2007 to 2019. He is now a Professor in the Department of Computer Science and Information Engineering, National Cheng Kung University, Taiwan. His research interests include digital content analysis, multimedia indexing, deep learning, and pattern recognition.

All faculties and students are welcome to join.