About
Jeong-hun Hong
Jeong-hun Hong is a researcher in the field of multimodal learning, integrating Computer Vision and Natural Language Processing. His expertise centers on multimodal representation learning and retrieval. Building upon this foundational work, his current research interests extend to Large Language/Multimodal Models and long video understanding, aiming to develop sophisticated and practical applications for comprehensive multimodal understanding and interaction.
| (+82)10-5148-6355 | invhun@gmail.com | GitHub Blog | Google Scholar | CV |
🎓 Education
M.S. | Department of Computer Science and Engineering (2023-2025)
- Kyungpook National University, Daegu, South Korea
- Thesis: Matryoshka Contrastive Learning via Embedding Dimension Rearrangement and Learnable Granularity-Wise Tempereatures for Vision-Language Models
- Advisor: Sang-hyo Park
B.S. | Department of Computer Science and Engineering (2017-2023)
- Kyungpook National University, Daegu, South Korea
📚 Publications
International conference
- Narrating the Video: Boosting Text-Video Retrieval via Comprehensive Utilization of Frame-Level Captions
Chan Hur, Jeong-hun Hong, Dong-hun Lee, Dabin Kang, Semin Myeong, Sang-hyo Park, and Hyeyoung Park Conference on Computer Vision and Pattern Recognition (CVPR), 2025. (*: Equal Contributors)
[paper | code]
International journal
- Research and Standards in 3D Scene Description Technologies: A Survey
Dong-shin Lim, Dong-hun Lee, Dong-hwi Kim, Jeong-hun Hong, Aro Kim, Chae-yeong Song, Bosung Baek, Dabin Kang, Myeong-jin Jang, and Sang-hyo Park APSIPA Transactions on Signal and Information Processing, SCIE, 2025. (Q2 Journal)
Domestic publications
Dataset Analysis and Comparative Evaluation for Fallen Person Detection in Virtual Environments
Hui-yan Yu, Jeong-hun Hong, Dabin Kang, and Sang-hyo Park Proceedings of the Korean Institute of Broadcast and Media Engineers Conference, 266-269, 2025
[paper]A Study on the Use of a Text-to-image Generative Model to Generate Consistent Story Illustrations
Semin Myeong, Dabin Kang, Chae-yeong Song, Jeong-hun Hong, and Sang-hyo Park Jounal of Broadcast Engineering, 29(6), 1043-1055, 2024.
[paper]Image Captioning Network for Disaster Dataset Based on Transfer Learning and Disaster Convolution Block
Jeong-hun Hong, Dong-hun Lee, Han-gyul Baek, Byungjun Bae, and Sang-hyo Park Journal of Korean Institute of Communications and Information Sciences (KICS), 48.8: 950-954, 2023.
[paper]
🌐 Standardization Activities
144th MPEG [detail]
[SD] Report on possible asset for scene description Jeong-hun Hong, Dong-hun Lee, Chae-yeong Song, Sang-hyo Park, Hyeon-Cheol Moon, Myeong-kyun Kim, Jinwoo Jeong, and Sungjei Kim ISO/IEC JTC 1/SC 29/WG 3 m64815, Oct. 2023, Hannover.
[SD] Guideline of format conversion from 3DSSG and 3RScan to glTF Jeong-hun Hong, Dong-hun Lee, Chae-yeong Song, and Sang-hyo Park ISO/IEC JTC 1/SC 29/WG 3 m65479, Oct. 2023, Hannover.
🏆 Honors & Awards
- CSE Thesis Award for Outstanding Graduate Student (2025) [view]
⚙️ Skills
- Machine Learning & Deep Learning: Pytorch, Tensorflow
- Knowledge: Multimodal learning, Video understanding, Vision-language model, Multimodal retrieval
- Language Skills: Korean, English