Home > 열린광장 > 열린광장 > 세미나공지
제목 [IT기술세미나]Overcoming the Curse of Data Scarcity in Modern Machine Learning(오태현박사-MIT CSAIL, Postdoctoral Associate)
작성자 수업담당 작성일 2018-09-06


9월 6() IT기술세미나 수업이 아래와 같이 진행됩니다. 관심 있는 학생들의 많은 참여 부탁드립니다.

(IT기술세미나 수업은 전공분야의 내용을 영어로 하는 수업이며, 대학원 수업이지만 수강생 여부와 상관없이 학부생도 참여가능하오니 관심 있는 모든 학생들이 참여해 주시기 바랍니다)

- 시간 : 오후 5~ 550

- 장소 : 1공학관 23219

- 세미나 연사 : 오태현박사( MIT CSAIL, Postdoctoral Associate)


Title: Overcoming the Curse of Data Scarcity in Modern Machine Learning

Abstract:
Recent progress in machine learning has led to many advances in engineering and science fields. Most notable successes have seen in supervised learning with deep neural networks. Despite these successes, as a consequence of deploying the high capacity models, a major limitation of the supervised approaches is to necessitate a massive amount of carefully annotated and curated data which is expensive.
In this talk, I will present four potential strategies to overcome data deficiency by highlighting a few application examples of them:
1) Generating synthetic data, 2) Putting a prior into architecture, 3) Using self-supervision and semi-supervision, and 4) Implanting into a classical algorithm. Since these are certainly not the only ones, I will discuss some promising directions to overcome the limited data.

Bio:
Tae-Hyun Oh is a postdoctoral associate at the Computer Science and Artificial Intelligence Laboratory (CSAIL), Massachusetts Institute of Technology (MIT), Cambridge, MA, USA. He received the B.E. degree (First class honors) in Computer Engineering from Kwang-Woon University, South Korea in 2010, and the M.S. and Ph.D. degrees in Electrical Engineering from KAIST, South Korea in 2012 and 2017, respectively. He was a research intern at Microsoft Research, Beijing, and Redmond in 2014 and 2016, respectively. He was awarded the Microsoft Research Asia fellowship, a gold award from Samsung HumanTech thesis, Qualcomm Innovation award, and the top research achievement award from KAIST. His research interests include machine perception understanding, machine learning and its computer vision and graphics applications.