Human Parsing: Traditional and Deep Models

题 目】Human Parsing: Traditional and Deep Models

主讲人】 Dr. Si Liu

Chinese Academy of Sciences

时 间】 2015年12月04日(周五)下午14:00

地 点】 北京大学 计算机所大楼 106报告厅

Abstract:In this talk, I shall introduce in detail our several recent works in the fashion parsing. The fashion parsing problem suffers from two bottlenecks. One is lack of training data, the other is missing an end-to-end fashion parser. To solve the problem of insufficient training samples, we propose a weakly supervised method, namely fashion parsing with weak color-category labels and a semi-supervised method called transferred human parsing with video context. To tackle the second challenge, we propose two deep learning based methods, including a parametric framework namely active template regression and a non-parametric framework called Matching-CNN.

Bio:Dr. Si Liu is now an Associate Professor in Institute of Information Engineering, Chinese Academy of Sciences. She used to be a Research Fellow at the Department of Electrical and Computer Engineering, National University of Singapore (NUS). She obtained PhD degree from Institute of Automation, Chinese Academy of Sciences (CASIA) in 2012. She obtained Bachelor degree from Experimental Class of Beijing Institute of Technology (BIT). Her current research interests include attribute prediction, object detection and image parsing. She is also interested in the applications, such as makeup and clothes recommendation, online product retrieval. She received the Best Paper Awards from ACM MM’13, Best Demo Awards from ACM MM’12.

CLOSE

上一篇 下一篇