Jaejun Hwang

I am an Algorithm enginner at GENGENAI, focusing diffusion model for image generation. Before joining GENGENAI, I was a graduate student in the Graduate School of Artificial Intelligence integrated M.S. program at POSTECH. I was a member of the Computer Vision Lab at POSTECH under Minsu Cho.
My research interests lie in computer vision and deep learning, especially in generation model and temporal consistency. If you are interested in my research projects, please feel free to contact me by clicking one of the icons below.

News

Jun 25, 2025	📝 Our paper ‘Generic Event Boundary Detection via Denoising Diffusion’ is accepted to ICCV 2025.
Apr 1, 2025	📝 I joined GENGENAI as an algorithm engineer.

Experience

April 2025 - Present	GENGENAI, Seoul, South Korea Algorithm Engineer
March 2022 – Feburary 2025	Computer Vision Lab, POSTECH, Pohang, South Korea M.S student Advisor: Minsu Cho

Education

March 2022 - Feburary 2025	Pohang University of Science and Technology (POSTECH), Pohang, South Korea M.S. Student in Graduate School of Artificial Intelligence
March 2015 - Feburary 2022	Hanyang University, Seoul, South Korea B.S in Department of Computer Science

Publications

Generic Event Boundary Detection via Denoising Diffusion

Jaejun Hwang*, Dayoung Gong*, Manjin Kim, and Minsu Cho

IEEE/CVF International Conference on Computer Vision (ICCV), 2025

Abs arXiv Website

Generic event boundary detection (GEBD) aims to identify natural boundaries in a video, segmenting it into distinct and meaningful chunks. Despite the inherent subjectivity of event boundaries, previous methods have focused on deterministic predictions, overlooking the diversity of plausible solutions. In this paper, we introduce a novel diffusion-based boundary detection model, dubbed DiffGEBD, that tackles the problem of GEBD from a generative perspective. The proposed model encodes relevant changes across adjacent frames via temporal self-similarity and then iteratively decodes random noise into plausible event boundaries being conditioned on the encoded features. Classifier-free guidance allows the degree of diversity to be controlled in denoising diffusion. In addition, we introduce a new evaluation metric to assess the quality of predictions considering both diversity and fidelity. Experiments show that our method achieves strong performance on two standard benchmarks, TAPOS and Kinetics-GEBD, generating diverse and plausible event boundaries.

Professional Services

Jan 1, 2024	Professional Service

Honors and Awards

Hanyang Brain (Merit-based) Scholarship, Hanyang CSE, 2019