Von Mises-Fisher Mixture Model-based Deep learning and its Application to Face Verification

讲座名称: Von Mises-Fisher Mixture Model-based Deep learning and its Application to Face Verification
讲座时间 2018-05-18
讲座地点 管理学院313会议室
讲座人 Liming Chen
讲座题目:Von Mises-Fisher Mixture Model-based Deep learning and its Application to Face Verification
讲座人:Prof. Liming Chen
讲座摘要:A number of pattern recognition tasks, e.g. , face verification, can be boiled down to classification or clustering of unit length directional feature vectors whose distance can be simply computed by their angle. In this paper, we propose the von Mises-Fisher (vMF) mixture model as the theoretical foundation for an effective deep-learning of such directional features and derive a novel vMF Mixture Loss and its corresponding vMF deep features. The proposed vMF features learning achieves a discriminative learning, i.e. , compacting the instances of the same class while increasing the distance of instances from different classes, and subsumes a number of loss functions or deep learning practice, e.g. , normalization. The experiments carried out on face verification using 4 different challenging face datasets, i.e. , LFW, IJB-A, YouTube faces and CACD, show the effectiveness of the proposed approach, which displays very competitive and state-of-the-art results.



Liming Chen is a Professor in the Department of Mathematics and Computer Science, Ecole Centrale de Lyon, University of Lyon, France. He received his BSc in Mathematics and Computer Science from the University of Nantes in 1984, his MSc and PhD in computer science from the University Pierre and Marie Curie Paris 6 in 1986 and 1989 respectively. He was an associate professor at the Université de Technologie de Compiègne before he joined Ecole Centrale de Lyon as Professor in 1998. He served as the Chief Scientific Officer in the Paris-based company Avivias from 2001 to 2003, and the scientific multimedia expert in France Telecom R&D China in 2005. He was the head of the Department of Mathematics and Computer science from 2007 through 2016. His current research interests include computer vision, machine learning, image and video analysis and categorization, face analysis and recognition, and affective computing. Liming has over 250 publications and successfully supervised over 35 PhD students. He has been a grant holder for a number of research grants from EU FP program, French research funding bodies and local government departments. Liming has so far guest-edited 2 journal special issues. He is an associate editor for Eurasip Journal on Image and Video Processing and a senior IEEE member.

讲座视频 暂无视频

(转载文章,请注明出处: 西安交通大学学术资源平台 http://meeting.xjtu.edu.cn)