Guisik Kim 

Senior Researcher 
Intelligent Image Processing Research Center
Korea Electronics Technology Institute, Seongnam, Korea

Ph.D.
School of Computer Science and Engineering, Chung-Ang University
Seoul, Korea

CV (231030 updated)

E-mail: specialre@naver.com / specialre@keti.re.kr
‌
‌
Major Interests
​- Dehazing
- Deepfake detection
- Low-level vision and its applications
- Satellite Image Deep Learning
‌- Scene understanding in adverse weather conditions


News

- Our paper was accepted at WACV 2025
- Our paper was accepted at the AIM ECCV workshop 2024
- Our paper has been accepted by GRSL
- I delivered an invited talk at the Early Career Researchers Session of the Korean Institute of Broadcast and Media Engineers, 2023
- Our paper was accepted at the NTIRE CVPR workshop 2023
- Recognized as an Outstanding Reviewer, ACCV 2022
- Successfully defended my Ph.D. thesis
- Our paper was accepted by TIP
- Awarded 1st place in the AI Graduate School Challenge 2022
- Received the EiR Award 2022 from the School of Computer Science and Engineering, CAU
- Our paper was accepted by T-ITS
- Our paper was accepted by BMVC 2020
- Our paper was accepted for an oral presentation at ECCV workshop 2020
- Achieved 4th place in the AIM Real-World Super-Resolution track
- Our paper was accepted for spotlight talks at the ICCV workshop 2019
- Our paper was accepted by CVPR workshop 2019
- Our paper was accepted by ICIP 2019
- Our paper was accepted by ICIP 2018
- Our paper was accepted by BMVC 2017

Publications

Journal Publications

Satellite Image Dehazing based on Dual Frequency Pass Networks

Guisik Kim, Choongsang Cho, Joohyung Kang, and Junseok Kwon, Satellite Image Dehazing based on Dual Frequency Pass Networks. IEEE Geoscience and Remote Sensing Letters, Accepted, 2024

temp.jpg

Dual Gradient based Snow Attentive Desnowing

Guisik Kim, Sungmin Cho, Dokyeong Kwon, Seo Hyeon Lee, and Junseok Kwon, Dual Gradient based Snow Attentive Desnowing. IEEE Access, Accepted, 2023
​

Self-Parameter Distillation Dehazing

Guisik Kim, and Junseok Kwon, Self-Parameter Distillation Dehazing. IEEE Transactions on Image Processing (TIP), Dec. 2022. (IF 10.454, SCI TOP 10%)
​

Attentive Style Transfer with Target Feature Palette

- Suhyeon Ha, Guisik Kim, and Junseok Kwon, Multimedia Tools and Applications (MTAP), accepted, 2022

- Suhyeon Ha, Guisik Kim, and Junseok Kwon, The 16th European Conference on Computer Vision (ECCV) Workshop Oral presentation, SEC, Glasgow, Scotland (Online), 2020. (Top 3 papers)
​

Deep Illumination-aware Dehazing with Low-light and Detail Enhancement

Guisik Kim, and Junseok Kwon, Deep Illumination-aware Dehazing with Low-light and Detail Enhancement. IEEE Transactions on Intelligent Transportation Systems (T-ITS), vol.23, 2022. (IF 6.492, SCI TOP 10%)

Pixel-wise Wasserstein Autoencoder for Highly Generative Dehazing

Guisik Kim, Sung Woo Park, and Junseok Kwon, Pixel-wise Wasserstein Autoencoder for Highly Generative Dehazing. IEEE Transactions on Image Processing (TIP), vol.30, page 5452 - 5462, 2021. (IF 10.454, SCI TOP 10%)
​
- 대한전자공학회 2021년도 추계학술대회 우수 저널 초청 강연
​
- 2022 인공지능대학원 (AIGS) 심포지엄 우수성과발표

Robust Person Re-identification via Graph Convolution Networks

- Guisik Kim, Dong Wook Shu, and Junseok Kwon, Multimedia Tools and Applications (MTAP), vol. 80, page 29129 - 29138, 2021

- Guisik Kim, Dong wook Shu, and Junseok Kwon, International Conference on Computer Vision (ICCV) Workshop Spotlight Talks, Seoul, Korea, 2019.


Visual Tracking with Adaptive Initial Configuration and Likelihood Landscape Analysis

Guisik Kim and Junseok Kwon, IET Computer Vision, vol. 13 , no. 1, Feb, page 1-7, 2019. 

Effective Visual Tracking using Multi-block and Scale Space based on Kernelized Correlation Filters

Soowoong Jeong, Guisik Kim, and Sangkeun Lee,  Sensors, vol. 17, no. 3, Feb.  2017.

Conference Publications

Satellite Image Dehazing Via Masked Image Modeling and Jigsaw Transformation

Guisik Kim, Choongsang Cho, and Junseok Kwon, Satellite Image Dehazing Via Masked Image Modeling and Jigsaw Transformation. Advances in Image Manipulation workshop in conjunction with ECCV, Accepted, 2024

illustration.png

Deep Dehazing Powered by Image Processing Network

Guisik Kim, Jinhee Park, and Junseok Kwon, New Trends in Image Restoration and Enhancement workshop and associated challenges in conjunction with CVPR 2023, Vancouber, Canada, 2023. 
overview.png

AIM 2022 Challenge on Instagram Filter Removal: Methods and Results

Guisik Kim, and Seohyeon Lee, et al., European Conference on Computer Vision (ECCV) Workshop,  Tel Aviv,  Israel, 2022. 

DALE : Dark Region-Aware Low-light Image Enhancement

Dokyeong Kwon*, Guisik Kim*, and Junseok Kwon, The 31th British Machine Vision Conference (BMVC), Manchester, UK (Online), 2020.
* equal contribution

Attentive Style Transfer with Target Feature Palette

- Suhyeon Ha, Guisik Kim, and Junseok Kwon, The 16th European Conference on Computer Vision (ECCV) Workshop Oral presentation, SEC, Glasgow, Scotland (Online), 2020. (Top 3 papers)
​
- Suhyeon Ha, Guisik Kim, and Junseok Kwon, Multimedia Tools and Applications (MTAP), accepted, 2022

AIM 2019 challenge on real-world image super-resolution: Methods and results

Guisik Kim, and Dokyeong Kwon, et al., International Conference on Computer Vision (ICCV) Workshop, Seoul, Korea, 2019. (4th place of the real-world image super-resolution challenge)

Robust Person Re-identification via Graph Convolution Networks

- Guisik Kim, Dong wook Shu, and Junseok Kwon, International Conference on Computer Vision (ICCV) Workshop Spotlight Talks, Seoul, Korea, 2019.

- Guisik Kim, Dong Wook Shu, and Junseok Kwon, Multimedia Tools and Applications (MTAP), vol. 80, page 29129 - 29138, 2021

Bidirectional Deep Residual learning for Haze Removal 

Guisik Kim, Jinhee Park, Suhyeon Ha and Junseok Kwon, Computer Vision and Pattern Recognition (CVPR) Workshop, Long beach, USA, 2019. 

Low-lightGAN: Low-Light Enhancement via Advanced Generative Adversarial Network with Task-Driven Training

Guisik Kim, Dokyeong Kwon, and Junseok Kwon, The 26th IEEE International Conference on Image Processing (ICIP), Taipei, Taiwan, 2019. 

LEDNet: Deep Illumination-aware Dehazing with Low-light Enhancement

Guisik Kim and Junseok Kwon, Workshop and Challenge on Perceptual Image Restoration and Manipulation in conjunction with ECCV 2018, Munich, Germany. 2018. (Short version)

Adaptive Patch based on Convolution Neural Network for Robust Dehazing

Guisik Kim, Suhyeon Ha, and Junseok Kwon, The 25th IEEE International Conference on Image Processing (ICIP), Athens, Greece. 2018.

Robust Pixel-wise Dehazing Algorithm based on Advanced Haze-Relevant Features

Guisik Kim and Junseok Kwon, The 28th British Machine Vision Conference (BMVC), London, UK. 2017.

Professional Activities 

Journal Reviewer: 
  1. IEEE Transactions on Pattern Analysis and Machine Intelligence 
  2. ​IEEE Transactions on Image Processing
  3. IEEE Signal Processing Letters
  4. Artificial Intelligence Review
  5. Neurocomputing
  6. Journal of Visual Communication and Image Representation
​Conference Reviewer:
  1. ​CVPR, ICCV, ECCV
  2. BMVC, AAAI
  3. ACCV
  4. NTIRE, AIM workshop   
​

Awards

2022. 12. Outstanding Reviewer Award, ACCV 2022
2022. 08. 1st place award in  AI Graduate School Challenge 2022: Minister of Sicence and ICT Award
2022. 07. EiR Award 2022 from the School of Computer Science and Engineering, CAU 
201‌9. 02. 3‌1th Image Processing and Image Understanding : Silver Prize
2018. 02. 30th Image Processing and Image Understanding : Silver Prize

Patents

Guisik Kim, Sung Woo Park, and Junseok Kwon. KOR Patent Pixel-wise Wasserstein Autoencoder for Highly Generative Dehazing (2021)
Guisik Kim and Junseok Kwon. KOR Patent 10-2020-0178743, Self-Distillation Dehazing (2020)
Guisik Kim, Dokyeong Kwon, and Junseok Kwon. KOR Patent 10-2020-0174683, DALE: Dark Region-Aware Low-light Image Enhancement (2020)
Guisik Kim, Dokyeong Kwon, and Junseok Kwon. KOR Patent 10-2019-0077222 (10-2134405-0000), System and Method for Improving Low Light Level Image Using Generative Adversarial Network (2019)