Home

I am currently an associate professor at the Department of Computer Science, Yonsei University.

My research focuses on multiple aspects of photography - from capturing images, processing and editing images, and to managing and interacting with photographs. The vision is to integrate different areas of computer science - computer vision, computer graphics, computational photography, and machine learning - to enhance multiple aspects of the photography culture.

More information can be found on my CV & Google Scholar page.

News

I am always looking for talented and motivated collaborators (graduate students, postdocs, and undergraduate interns).

Feel free to contact me through email (seonjookim(at)yonsei.ac.kr), if you're interested in working with me.

September 2018

Our work on TAGAN (Text Adaptive Generative Adversarial Network) is accepted in NIPS 2018. Our first paper in NIPS!

August 2018

Our baseball database (BBDB) is available online!

https://sites.google.com/site/eccv2018bbdb/

July 2018

Team from our lab (Chang Ha, Gunhee, Dasol, Hyunjin, In, Woongoh) won the AI R&D Challenge hosted by the Ministry of Science and ICT. Congratulations!

July 2018

A paper on Fine-grained Video Dataset is accepted in ECCV 2018.

May 2018

Seoung Wug is once again off to Silicon Valley for an internship at Adobe Research.

Seounghyeon is off to Los Angeles for an internship at Snapchat Research.

March. 2018

We have 3 papers accepted for CVPR 2018!

A paper on color transfer is accepted in CVIU.

Feb. 2018

Our deep network "EPINET" is currently #1 in the light field depth estimation benchmark:

http://hci-lightfield.iwr.uni-heidelberg.de/benchmark/table?column-type=images&metric=badpix_0070

Student Younghyun awarded with the Bronze prize in the Samsung Humantech Paper Award.

I will serve as an Area Editor for CVIU.

Aug 2017

I will serve as an Area Chair and Demo Chair for CVPR 2018.

Seoung Wug is off to Silicon Valley for an internship at Adobe.

July 2017

A paper on modelling the scene dependent imaging in cameras using a DNN is accepted for ICCV 2017.