Dong-A University Doctoral Student Yoho Lee Selected for Outstanding R&D Achievement in Busan

In the 'Outstanding Academic' Category...
BISTEP President's Award and 1 Million Won Prize

The achievement of a research team at Dong-A University, which developed deep learning technology capable of restoring low-resolution videos to high-quality, has been selected as an outstanding research and development achievement in the Busan region.

Dong-A University (President Lee Hee-woo) announced on September 28 that Yoho Lee, a doctoral student in Professor Dongsan Jeon's Video Signal Processing Lab within the Department of Computer Engineering at the Graduate School, was ultimately selected as one of the 'Outstanding Research and Development Achievements of the Busan Region 2026.'

This project, organized by Busan Metropolitan City and the Busan Institute for Science, Technology and Education Promotion (BISTEP), aims to identify and promote exemplary research and development results within the region. Lee was recognized in the 'Outstanding Academic' category, which targets graduate students and early-career researchers, receiving the BISTEP President's Award and a prize of 1 million won.

The selected research achievement is a deep learning-based 'video super-resolution technology' that restores low-resolution and low-quality videos to high resolution and high quality.

While previous studies have focused on preprocessing stages such as aligning and combining multiple frames, doctoral student Yoho Lee concentrated on the restoration phase that refines the final output. The research team newly proposed a purification module (USTM) dedicated to the restoration process, pioneering a relatively unexplored area in earlier studies.

The related research paper was published in 'IEEE T-CSVT,' an international journal in the field of electrical and electronic engineering. The journal has a citation index of 10.8 and ranks in the top 3.4% of journals according to JCR.

A key feature of this technology is that it is designed so that it can be added as a component to existing video quality enhancement systems without requiring a complete replacement. Consistent quality improvement effects were confirmed across three standard evaluation datasets and four different video restoration models, and by using fine-tuning instead of full retraining, the number of training iterations required was reduced by an average of 73%, thereby lowering computational resource and power consumption.

In particular, the research team expects that this technology can also be used for the restoration and advancement of video assets owned by Busan, a UNESCO Creative City of Film.

They explain that, rather than artificially generating information not present in the original, the system is designed to improve image quality while preserving the original content. This makes it applicable not only to restoring past movies and video films with low or damaged resolution, but also to various fields that require high-resolution images such as UHD broadcasting, OTT services, autonomous driving, and medical imaging.

Doctoral student Yoho Lee said, "I am deeply honored that our achievement has been recognized not only for its academic contribution, but also for its potential for industrial expansion in the region," adding, "We will strive to develop technology that can be practically used in the field of video signal processing and AI-based video restoration."

Professor Dongsan Jeon, Lee’s advisor, stated, "This achievement is the result of the student's consistent effort combined with our department's research-centered educational environment," emphasizing, "We will continue providing systematic support so that local young talents can grow into highly competitive research professionals at the international level."

The Video Signal Processing Lab, led by Professor Dongsan Jeon, focuses on video signal processing, computer vision, AI-based video restoration, and 3D scene representation and rendering based on 3D Gaussian Splatting. Currently, the lab has 10 personnel, including master’s and doctoral students as well as undergraduate interns, and is committed to nurturing research talent.

Dr. Yoho Lee, a doctoral candidate in the Department of Computer Engineering at Dong-A University (left), and Professor Dongsan Jeon, his advisor.

This content was produced with the assistance of AI translation services.