833. Deep Weakly Supervised Learning in Visual Recognition

 연사: 한보형 서울대학교 전기정보공학부 교수
 일시: 2018년 03월 15일(목), 17:00 ~ 18:30
 장소: 서울대학교 제1공학관(301동) 118호

Abstract: The success of deep learning in computer vision is partly attributed to the construction of large-scale labeled datasets such as ImageNet. However, computer vision problems often require substantial human efforts and interventions to obtain accurate annotations due to dynamic aspects of class labels, needs for pixel-level labeling, and annotation ambiguities. Hence, collecting high quality large-scale labeled datasets is very time consuming and even infeasible. This talk mainly addresses challenges and approaches related to data deficiency issue in the context of computer vision including object detection, semantic segmentation, action localization, and many others, where we can take advantage of various types of weakly supervision.

Biography:
 Education
2005 PhD, the Department of Computer Science at the University of Maryland, College Park, USA
2000 MS, the Department of Computer Engineering at Seoul National University, Korea
1997 BS, the Department of Computer Engineering at Seoul National University, Korea
 Career
Tutorial Chair in ICCV 2019
Associate Professor in the Department of Electrical and Computer Engineering at Seoul National University
Demo Chair in ACCV 2014
Associate Professor in the Department of Computer Science and Engineering at POSTECH
Visiting research scientist in Machine Intelligence Group at Google, Venice, CA, USA
Research scientist in Mobileye Vision Technologies, Princeton, NJ, USA and Samsung, Irvine, CA, USA

 Major Research Topics
Computer vision, Image understanding, Machine learning