MCL에서 2022년 3월부터 아래 분야를 연구할 대학원생을 모집합니다. 관심 있는 사람은 김창수 교수(changsukim@korea.ac.kr)에게 연락하세요.

1)시작:  2022년 3월 (2021년 9월도 가능)

2) TO:   3명

3) 연구분야: Machine Learning, Computer Vision, Image Processing

Our techniques on interactive video segmentation and semantic line detection will be presented in CVPR 2021.

• * Yuk Heo, Yeong Jun Koh, and Chang-Su Kim, “Guided Interactive Video Object Segmentation Using Reliability-Based Attention Maps” (Oral)
• * Dongkwon Jin, Wonhui Park, Seong-Gyun Jeong, and Chang-Su Kim, “Harmonious Semantic Line Detection via Maximal Weight Clique Selection”

Congratulations, Yuk, for enduring his advisor’s transfer threats. Stay in KU, although he is unlikely to stop them.
Congratulations, Dongkwon, for becoming a research specialist with a lucky TEPS score.
Congratulations, Wonhui, for doing so many things in his first year in MCL.
It is OK, the others! So far we have had 22 CVPR submissions rejected. Many of them were good papers and have been published eventually. Don’t let this frustrate you.
We appreciate MCL alumni Dr. Koh and Dr. Jeong for their excellent supervision of younger members.

We are pleased to announce that another order learning paper will be presented in ICLR.

Seon-Ho Lee and Chang-Su Kim, “Deep Repulsive Clustering of Ordered Data Based on Order-Identity Decomposition,” ICLR 2021.

Congratulations to Seon-Ho for his first top-tier conference paper!

This year, we will present our techniques on video interpolation, depth estimation, semantic line detection, image enhancement, and interactive video segmentation in ECCV2020.

* Junheum Park, Keunsoo Ko, Chul Lee, Chang-Su Kim, “BMBC:Bilateral Motion Estimation with Bilateral Cost Volume for Video Interpolation”
* Jae-Han Lee, Chang-Su Kim, “Multi-Loss Rebalancing Algorithm for Monocular Depth Estimation”
* Dongkwon Jin, Jun-Tae Lee, Chang-Su Kim, “Semantic Line Detection Using Mirror Attention and Comparative Ranking and Matching”
* Han-Ul Kim, Yeong Jun Koh, Chang-Su Kim, “Global and Local Enhancement Networks For Paired and Unpaired Image Enhancement”
* Han-Ul Kim, Yeong Jun Koh, Chang-Su Kim, “PieNet: Personalized Image Enhancement Network”
* Yuk Heo, Yeong Jun Koh, Chang-Su Kim, “Interactive Video Object Segmentation Using Global and Local Transfer Modules”

Congratulations, Junheum, Keunsoo, Jae-Han, Dongkwon, Han-Ul, and Yuk!
Many thanks to our alumni Prof. Chul Lee, Prof. Young Jun Koh, Dr. Jun-Tae Lee for their excellent guidance of younger MCL members.