cv
General Information
Full Name | Saehoon Kim |
Languages | Korean, English(working proficiency) |
Education
-
2018 Ph.D. in Computer Science and Engineering
POSTECH, Pohang, South Korea - Supervisor: Prof. Seungjin Choi
- Dissertation: Binary Embedding: Theory and Applications
-
2009 B.S. in Computer Science and Engineering
POSTECH, Pohang, South Korea
Experience
-
2020 - 2024 Director, Visual generative modeling
Kakao Brain - Leading the Karlo team in commercializing T2I solutions.
- Some of our products are available at karlo.ai and karlo-profile.
-
2017 - 2020 Research Director
AITRICS - Leading the VitalCare focused on early prediction of sepsis.
- Managing a machine learning research team of ten members.
-
2013 Research Intern
Microsoft Research, Redmond, US. - Developing a predictive parallelization technique to reduce tail latency in web search systems, published at WSDM-2015.
- This project was advised by Yuxiong He and Sameh Elnikety.
-
2012 Research Intern
Microsoft Research, Beijing, China. - Developing a scalable solution for billion-scale near-duplicate image discovery, published at WACV-2015.
- This project was advised by Xin-Jing Wang.
Honors and Awards
-
2016 - NAVER Ph.D. Fellowship
-
2015 - WSDM’15 Best Paper Runner-up Award
- POSTECH-Qualcomm Fellowship
- Korea BI Data Mining Conference Best Paper Award
-
2012 - Korea BI Data Mining Conference Best Paper Award
-
2002 - Bronze Prize, Korea Olympiad in Informatics
Academic Interests
-
Generative Models for Healthcare
- CXR report generation solution built upon multi-modal language models.
- Predictive models for major clinical events in ICU settings.
-
Generative Models for Computer Vision
- Diffusion models for conditional image generation.
- Efficient VQ methods for autoregressive models in image generation.