Faculty Research Profile
Jaejun Yoo
학력
• 2018: Ph.D. Bio and Brain Engineering, KAIST
• 2013: M.S. Bio and Brain Engineering, KAIST
• 2011: B.S. Bio and Brain Engineering, KAIST
주요 경력
• 2021 - Present: Assistant Professor, UNIST
• 2019 - 2021: Post-doc, École polytechnique fédérale de Lausanne (EPFL)
• 2018 - 2019: Research Scientist, Clova AI lab, NAVER
수상/학회/외부활동
• 2018 Samsung Human Tech. Award
• NeurIPS'20 Best Reviewer Award (Top 10% among 7062 reviewers).
• 2023 - 현재 대한의료인공지능학회 (Korea Society of Artificial Intelligence in Medicine) 총무이사
• 2023- 현재 IEEE SPS Computational Imaging Technical Committee (CI TC member).
컴퓨터 비전 및 바이오 영상신호처리 연구실
Laboratory of Advanced Imaging Technology (LAIT)
컴퓨터 비전 및 바이오 영상신호처리 연구실에서는 사람과 영상 사이의 다양한 상호작용을 보다 더 쉽고, 정확하고, 직관적으로 만들기 위한 모델을 연구합니다. 이를 위해 새로운 기계학습 기법을 개발하되, 단순히 기능을 하는 모델을 넘어서서 신호처리 기반의 분석을 바탕으로 효율적이고, 성능이 좋으면서도, 동작 원리를 잘 이해할 수 있는 모델을 만드는 것을 목표로 합니다.
Our main research area lies at the intersection of computer vision, machine learning, and inverse problems, including natural image recovery and medical imaging. We have a strong interest in generative models, representation learning, and the use of signal processing theories for image processing. Our goal is to build strong and intelligent signal processing models, capable of recreating the world we perceive. We aim to establish a bridge between signal processing and deep learning, taking the best of both worlds. We study deep learning models that learn structural priors by synthesizing and modeling millions of images and videos. Along the way, the learned models provide insights to seek mathematical elegance and a clear understanding of our world, which in turn encourages us to find better models to analyze signals in nature.
Our main research area lies at the intersection of computer vision, machine learning, and inverse problems, including natural image recovery and medical imaging. We have a strong interest in generative models, representation learning, and the use of signal processing theories for image processing. Our goal is to build strong and intelligent signal processing models, capable of recreating the world we perceive. We aim to establish a bridge between signal processing and deep learning, taking the best of both worlds. We study deep learning models that learn structural priors by synthesizing and modeling millions of images and videos. Along the way, the learned models provide insights to seek mathematical elegance and a clear understanding of our world, which in turn encourages us to find better models to analyze signals in nature.
연구분야
생성 AI, 계산 이미징, 바이오 의료 영상, 신호처리 / generative models, computational imaging, biomedical imaging, signal processing
generative models, computational imaging, biomedical imaging, signal processing
연구주제
Inverse problems for various imaging modalities:
- natural image restorations (super-resolution, denoising, deblurring, etc.)
- medical image reconstructions (MRI, CT, SIM, Cryo-EM, DOT, EEG, fMRI, etc.)
Bridging between signal processing and deep learning communities:
- providing a design principle for deep learning architectures
- network analysis using topological data analysis (TDA)
Deep generative models:
- developing a high fidelity and diverse image-to-image translation model
- improving generative models based on theoretical understandings
Inverse problems for various imaging modalities:
- Natural image restorations (super-resolution, denoising, deblurring, etc.)
- Medical image reconstructions (MRI, CT, SIM, Cryo-EM, DOT, EEG, fMRI, etc.)
Bridging between signal processing and deep learning communities:
- Providing a design principle for deep learning architectures
- Network analysis using topological data analysis (TDA)
Deep generative models:
- Developing a high fidelity and diverse image-to-image translation model
- Improving generative models based on theoretical understandings
국가연구개발사업 기술 분류체계
국가과학기술표준분류
EE. 정보/통신 > EE01. 정보이론 > EE0108. 인공지능
논문
•ICLR / STREAM: Spatio-TempoRal Evaluation and Analysis Metric for Video Generative Models/Pum Jun Kim, Seojun Kim, Jaejun Yoo/2024
•AAAI / Can We Find Strong Lottery Tickets in Generative Models?/Sangyeop Yeo, Yoojin Jang, Jy-yong Sohn, Dongyoon Han, Jaeju
•TMI / Time-Dependent Deep Image Prior / J. Yoo, K.H. Jin, H. Gupta, J. Yerly, M. Stuber, M. Unser / 2021
•CVPR / StarGAN v2: Diverse Image Synthesis for Multiple Domains / Y.J. Choi, Y.J. Uh, J.Yoo, J.W. Ha / 2020
•CVPR / Rethinking Data Augmentation for Image Super-resolution: A Comprehensive Analysis and a New Strategy / J. Yoo, N.H. Ahn, K.A. Sohn / 2020
•ICCV / Photorealistic Style Transfer via Wavelet Transforms / J.Yoo, S.H. Chun, Y.J. Uh, B. Kang, J.W. Ha / 2019
•ICLR / Large-Scale Answerer in Questioner’s Mind for Visual Dialog Question Generation/ S.W. Lee, T. Gao, S. Yang, J. Yoo, J.W. Ha / 2019