Blog - Hyungjin

Welcome to my personal homepage. I am an ongoing Ph.D. student @KAIST. The motivation of my research is “Making invisible, visible”. Mostly, I work on solving inverse problems arising in computational imaging, and I focus on using generative modeling, especially diffusion models. I am both interested in generally advancing diffusion model-based inverse problem solvers, and applying them to solve important real-world problems.

Hyungjin Chung

Ph.D. student - Deep Learning & Inverse Problems

Korea Advanced Institute of Science and Technology (KAIST) BISPL

Biography

Second year Ph.D. student @KAIST bio-imaging signal processing & learning lab (BISPL). Prior research intern at the Los Alamos National Laboratory (LANL) applied math and plasma physics group (T-5). Diffusion models and inverse problems enthusiast. Hyungjin Chung has pioneered and advanced some of the most widely acknowledged works on diffusion model-based inverse problem solvers. Interested in 1) Advancing and widening the applicability of diffusion models in inverse imaging, 2) Acceleration of diffusion models, 3) Application to solve real-world problems (e.g. medical imaging).

Download my CV.

Interests

Deep Learning
Diffusion Model
Energy-based Model
Computational Imaging
Inverse Problems

Education

MS in Bio & Brain Engineering, 2021
Korea Advanced Institute of Science & Technology (KAIST)
BS in Biomedical Engineering, 2019
Korea University

Projects

Probabilistic Machine Learning - Kevin Murphy

News

DPS receives Gold award in 29th Samsung Humantech Paper Award (1st in Signal Processing)

Jan 16, 2023 12:00 AM

1 Paper accepted to ICLR 2023 (spotlight)

Jan 15, 2023 12:00 AM

2 Papers accepted to NeurIPS 2022 / NeurIPS 2022 SBM Workshop

Sep 15, 2022 12:00 AM

1 Paper accepted to CVPR 2022

Mar 15, 2022 12:00 AM

Interview at BRIC

Apr 27, 2021 12:00 AM

See all events

Featured Publications

Hyungjin Chung, Jeongsol Kim, Michael T. Mccann, Marc L. Klasky, Jong Chul Ye

January 2023 ICLR 2023 (SPOTLIGHT)

Diffusion Posterior Sampling for General Noisy Inverse Problems

Diffusion posterior sampling enables solving arbitrary noisy (e.g. Gaussian, Poisson) inverse problems that are both linear or non-linear.

Hyungjin Chung, Jeongsol Kim, Sehui Kim, Jong Chul Ye

November 2022 In Arxiv

Parallel Diffusion Models of Operator and Image for Blind Inverse Problems

We propose a method to perform posterior sampling with diffusion models on blind inverse problems.

Hyungjin Chung, Dohoon Ryu, Michael T. Mccann, Marc L. Klasky, Jong Chul Ye

November 2022 In Arxiv

Solving 3D Inverse Problems using Pre-trained 2D Diffusion Models

We propose a method that can solve 3D inverse problems in the medical imaging domain using only the pre-trained 2D diffusion model augmented with the conventional model-based prior.

Hyungjin Chung, Byeongsu Sim, Dohoon Ryu, Jong Chul Ye

September 2022 In NeurIPS 2022

Improving Diffusion Models for Inverse Problems using Manifold Constraints

Manifold constraint dramatically improves the performance of unsupervised inverse problem solving using diffusion models.

Hyungjin Chung, Byeongsu Sim, Jong Chul Ye

March 2022 In CVPR 2022

Come-Closer-Diffuse-Faster Accelerating Conditional Diffusion Models for Inverse Problems through Stochastic Contraction

Come-close to diffuse-fast when solving inverse problems with diffusion models. We establish state-of-the-art results with only 20 diffusion steps across various tasks including SR, inpainting, and CS-MRI

Hyungjin Chung, Jong Chul Ye

October 2021 In MedIA

Score-based diffusion models for accelerated MRI

Score-based diffusion models beat supervised learning methods on MRI reconstruction.

Yoseob Han, Jaeduck Jang, Eunju Cha, Junho Lee, Hyungjin Chung, Myoungho Jeong, Tae-Gon Kim, Byeong Gyu Chae, Hee Goo Kim, Shinae Jun, Sungwoo Hwang, Eunha Lee, Jong Chul Ye

February 2021 In Nat. Mach. Int.

Deep learning STEM-EDX tomography of nanocrystals

Resolution of STEM-EDX tomography is greatly enhanced through unsupervised deep learning. We verified the method by comparing the optical properties.

Recent Publications

Hyungjin Chung, Jeongsol Kim, Michael T. Mccann, Marc L. Klasky, Jong Chul Ye (2023). Diffusion Posterior Sampling for General Noisy Inverse Problems. ICLR 2023 (SPOTLIGHT).

PDF Cite Code

Hyungjin Chung, Jeongsol Kim, Sehui Kim, Jong Chul Ye (2022). Parallel Diffusion Models of Operator and Image for Blind Inverse Problems. In Arxiv.

PDF Cite

Hyungjin Chung, Dohoon Ryu, Michael T. Mccann, Marc L. Klasky, Jong Chul Ye (2022). Solving 3D Inverse Problems using Pre-trained 2D Diffusion Models. In Arxiv.

PDF Cite

Hyungjin Chung, Byeongsu Sim, Dohoon Ryu, Jong Chul Ye (2022). Improving Diffusion Models for Inverse Problems using Manifold Constraints. In NeurIPS 2022.

PDF Cite Code

Eunju Cha, Hyungjin Chung, Jaeduck Jang, Junho Lee, Eunha Lee, Jong Chul Ye (2022). Low-dose sparse-view HAADF-STEM-EDX tomography of nanocrystals using unsupervised deep learning. In ACS Nano.

Cite

Hyungjin Chung, Byeongsu Sim, Jong Chul Ye (2022). Come-Closer-Diffuse-Faster Accelerating Conditional Diffusion Models for Inverse Problems through Stochastic Contraction. In CVPR 2022.

PDF Cite

Joon Yeul Nam, Hyungjin Chung, Kyu Sung Choi, Hyuk Lee, Seung Jun Han, Tae Jun Kim, Hosim Soh, Eun Kang, Soo_jeong Cho, Jong Chul Ye, Jong Pil Im, Sang Gyun Kim, Yoon Jun Kim, Joo Sung Kim, Jung-Hwan Yoo, Hyunsoo Chung, Jeong-Hoon Lee (2021). Deep Learning Model for Diagnosing Gastric Mucosal Lesions Using Endoscopic Images: Development, Validation, and Method Comparison. GIE.

PDF Cite

Hyungjin Chung, Jong Chul Ye (2021). Score-based diffusion models for accelerated MRI. In MedIA.

PDF Cite Code

Mehmet Akcakaya, Burhaneddin Yaman, Hyungjin Chung, Jong Chul Ye (2021). Unsupervised Deep Learning Methods for Biological Image Reconstruction and Enhancement. IEEE SPM.

PDF Cite

Yoseob Han, Jaeduck Jang, Eunju Cha, Junho Lee, Hyungjin Chung, Myoungho Jeong, Tae-Gon Kim, Byeong Gyu Chae, Hee Goo Kim, Shinae Jun, Sungwoo Hwang, Eunha Lee, Jong Chul Ye (2021). Deep learning STEM-EDX tomography of nanocrystals. In Nat. Mach. Int..

PDF Cite Code DOI

Gyutaek Oh, Byeongsu Sim, Hyungjin Chung, Leonard Sunwoo, Jong Chul Ye (2020). Unpaired deep learning for accelerated MRI using optimal transport driven cycleGAN. IEEE TCI.

PDF Cite

Eunju Cha, Hyungjin Chung, Eung Yeop Kim, Jong Chul Ye (2020). Unpaired training of deep learning tMRA for flexible spatio-temporal resolution. IEEE TMI.

PDF Cite

Hyungjin Chung, Jong Chul Ye (2020). Feature Disentanglement in generating three-dimensional structure from two-dimensional slice with sliceGAN. In Nat. Mach. Int..

PDF Cite Code

Hyungjin Chung, Jaehyun Kim, Jeong Hee Yoon, Jeong Min Lee, Jong Chul Ye (2020). Simultaneous super-resolution and motion artifact removal in diffusion-weighted MRI using unsupervised deep learning. In ArXiv.

PDF Cite

Hyungjin Chung, Eunju Cha, Leonard Sunwoo, Jong Chul Ye (2020). Two-Stage Deep Learning for Accelerated 3D Time-of-Flight MRA without Matched Training Data. In Medical Image Analysis.

PDF Cite

Joonhyung Lee, Hyunjong Kim, Hyungjin Chung, Jong Chul Ye (2020). Deep Learning Fast MRI Using Channel Attention in Magnitude Domain. IEEE ISBI.