I have mainly researched generative models with multi-modal data for vairous applications with different conditions.
Generative Models. e.g., Diffusion Models, Image-to-Image Translation, Style Transfer etc.
Multimodal Learning. e.g., Audio-guided Image Manipulation, Text-guided Image Manipulation, etc.
Meanwhile, I am mainly interested in instructive, interactive, and personalized practical applications by understanding the intention of user actions. Additionally, I am also interested in Generative models & understanding and learning relations or representations in images and between images and other modalities, visual perception(detection, detection), and their further extensions to the 3D or video applications.
-
Korea University, Seoul, Korea
- M.S. Student in Computer Science and Engineering
- Mar. 2021 - Aug. 2023
-
Dongguk University, Seoul, Korea
- B.S. in Computer Engineering
- Mar. 2014 - Aug. 2020
-
Naver AI Research Intern (Naver, Seongnam, Korea)
- June. 2022 - Nov. 2022
- Mentor: Gayoung Lee
-
Undergraduate Intern (Korea University CVLAB, Seoul, Korea)
- Aug. 2020 - Feb. 2021
- Advisor: Prof. Seungryong Kim
Controllable Style Transfer via Test-time Training of Implicit Neural Representation
Sunwoo Kim*, Youngjo Min, Younghoon Jeong and Seungryong Kim
ArXiv Preprint, 2022.
[Project Page] [arXiv]
DiffMatch: Diffusion Model for Dense Matching
Jisu Nam, Gyuseong Lee, Sunwoo Kim, Hyunsu Kim, Hyungwon Cho, Seyeon Kim, Seungryong Kim
International Conference on Learning Representations (ICLR, 1.2% Oral presentation), 2024.
[arXiv]
User-friendly Image Editing with Minimal Text Input: Leveraging Captioning & Injection Technique
Sunwoo Kim*, Wooseok Jang, Hyunsu Kim, Junho Kim, Yunjei Cho, Seungryong Kim and Gayoung Lee
[arXiv]
LANIT: LAnguage-Driven Unsupervised Image-to-Image Translation for Unlabeled Data
Jihye Park*, Sunwoo Kim*, Soohyun Kim*, Seokju Cho, Jaejun Yoo, Youngjung Uh and Seungryong Kim
Conference on Computer Vision and Pattern Recognition (CVPR), 2023.
[Project Page] [arXiv]
Deep Translation Prior: Test-time Training for Photorealistic Style Transfer
Sunwoo Kim*, Soohyun Kim* and Seungryong Kim
36th AAAI Conference on Artificial Intelligence, (AAAI) 2022.
[Github] [arXiv]
- Human talking face Generation Projects
- June. 2022 - Nov. 2022