About me

 Kyong Hwan, Jin


I'm an enthusiastic researcher on imaging science - always curious on imaging principles and background theory.  In image processing, my research domain ranges from designing algorithm to implementing software. Recent interests are image quality enhancement and real inverse problems from computational photography using learnable non-linear systems, a.k.a, deep neural networks. I believe that our research trials make a step toward the better human life.



* Experience


- Assistant Professor, Dept. of Information and Communication Engineering, DGIST, South Korea, Feb. 2021 ~


- Staff Engineer, Samsung Research, Camera T/F - Visual Technology Team, South Korea, Sep. 2019 ~ Feb. 2021


- Post-doc., Swiss Federal Institute of Technology in Lausanne (EPFL), Switzerland, Jun. 2016 - Aug. 2019

PI : Michael Unser Biomedical Imaging Group
Research project : sparse-view differentiated phase contrast CT using low-rank Fourier interpolator

- Post-doc., Dept. of Bio & Brain Engineering, KAIST, South Korea, Mar. 2015 - May. 2016

PI : Jong Chul Ye @ Bio-Imaging Signal Processing Lab.

Research project : sparse + low-rank matrix completion for various applications (impulse noise removal), deep learning approach for accelerated MRI


* Education
- Ph.D. , Dept. of Bio & Brain Engineering, KAIST, South Korea, Feb. 2015.
- B.S.   , Dept. of Bio & Brain Engineering, KAIST, South Korea, Feb. 2008.


* Awards

2019 IEEE SPS Best Paper Award for the noted paper : 

 ”Deep Convolutional Neural Network for Inverse Problems in Imaging,” 

IEEE Transactions on Image Processing, Volume 26, No. 9, September 2017

Grant on EPFL Fellows co-funded by Marie Sklodowska-Curie (2015 call, European Union’sHorizon 2020)

Samsung Humantech Paper Award - Silver Medal (2015, South Korea),

- Samsung Humantech Paper Award - Participation Prize (2004, South Korea)

The Presidential Science Scholarship (2004-2008, South Korea)


Research Summary

- Deep-learning based inverse problems in image processing

- Image Enhancement - denoising, HDR, interpolation, dealiasing on sub Nyquist sampling

- Sampling & Signal processing

- Reproducing Kernel Hilbert Space

- Compressed Sensing

MRI, Ultrasound, CT, Microscopy

- Image processing for Lossless/Lossy Image artifacts