Publications | CV | Google Scholar

Affiliation: Computer Vision TU, AI Research Center, SAIT, Samsung Electronics
Email: sonsang35(at)gmail(dot)com

I received my Ph.D. degree from the Department of ECE, Seoul National University, Seoul, Korea.
Computer Vision Lab, Advisor: Prof. Kyoung Mu Lee
Thesis: Generalized Resampling Model for Practical Image Super-Resolution

Research Interests

I am interested in deep learning and low-level image restoration problems, especially image super-resolution and deblurring. Recently I mainly working on reconstructing high-resolution images from real-world inputs. My research topics also include network optimization and acceleration to handle practical issues in deep learning.

News

• Now I am with Samsung Electronics!
  Computer Vision TU, AI Research Center, SAIT  

• I won the Youlchon AI Stars Scholarship 2022!  

• I won the KCCV Sang-Uk Lee Prize from KCCV 2022!  

• One paper got accepted to WACV 2024!
  • Reyhaneh Neshatavar, Mohsen Yavartanoo, Sanghyun Son, and Kyoung Mu Lee, “ICF-SRSR: Invertible scale-Conditional Function for Self-Supervised Real-world Single Image Super-Resolution.”  
• One paper got accepted to ICCV 2023!
  • Joonkyu Park, Sanghyun Son, and Kyoung Mu Lee, “Content-Aware Local GAN for Photo-Realistic Super-Resolution.”  
• Two papers got accepted to CVPR 2022!
  • Wooseok Lee, Sanghyun Son, and Kyoung Mu Lee, “AP-BSN: Self-Supervised Denoising for Real-World Images via Asymmetric PD and Blind-Spot Network.”
  • Reyhaneh Neshatavar, Mohsen Yavartanoo, Sanghyun Son, and Kyoung Mu Lee, “CVF-SID: Cyclic Multi-Variate Function for Self-Supervised Image Denoising by Disentangling Noise from Image.”  
• One paper got accepted to ICLR 2022!
  • Seungjun Nah, Sanghyun Son, Jaerin Lee, and Kyoung Mu Lee, “Clean Images are Hard to Reblur: Exploiting the Ill-Posed Inverse Task for Dynamic Scene Deblurring.”
  • Code and demo will be released soon.  

• I won the Qualcomm Innovation Fellowship Korea 2021!  

• One paper got accepted to ICCV 2021!
  • Geonwoon Jang, Wooseok Lee, Sanghyun Son, and Kyoung Mu Lee, “C2N: Practical Generative Noise Modeling for Real-World Denoising.”
  • Code and demo will be released soon.  
• Our paper got accepted to IEEE TPAMI (IF 16.39)!
  • Sanghyun Son, Jaeha Kim, Wei-Sheng Lai, Ming-Hsuan Yang, and Kyoung Mu Lee, “Toward Real-World Super-Resolution via Adaptive Downsampling Models.”
  • You can find the full code from our github.  
• One paper got accepted to CVPR 2021!
  • Sanghyun Son and Kyoung Mu Lee, “SRWarp: Generalized Image Super-Resolution under Arbitrary Transformation.”
  • You can find the demo code from our github.  
• I am co-organizing the 6th NTIRE workshop and challenges in conjunction with CVPR 2021. Me, Seungjun Nah and Suyoung Lee are in charge of the following image deblurring and video super-resolution challenge tracks:
  • Image Deblurring Track 1. Low Resolution, Track 2. JPEG artifacts,
  • Video Super-Resolution Track 1. Spatial, Track 2. Spatio-Temporal.
  • Detailed challenge reports will be presented in the upcoming NTIRE 2021 workshop.  
• I have implemented an efficient bicubic resizing algorithm in PyTorch that is compatible with the MATLAB reference. Please check it from here.

Awards and Honors

• The KCCV Sang-Uk Lee prize (Test of Time award) from KCCV 2022.
• Winner of Qualcomm Innovation Fellowship Korea 2021.
• Highly cited paper award: Department of ECE, Seoul National University, 2018.
• Challenge winner & Workshop best paper award: NTIRE 2017 Challenge on Single Image Super-Resolution, 2017.
• Graduated with summa cum laude: B.S. in Department of ECE, Seoul National University, 2017.

Personal Pages

• I enjoy playing video games, especially Nintendo Switch and some simulation games on Steam.
  • Gaming archive