Donghwan Lee
Integrated M.S. & Ph.D. Student
SNU AIBL Lab, advised by Prof. Kyungsu Kim
I work on generative diffusion models — particularly object removal and inpainting — and on carrying these methods into medical imaging. My recent focus is training-free and test-time approaches that lean on the priors already inside large generative models, so that editing and removal can work without task-specific datasets.
I am also interested in applying generative modeling to MRI, CT, and X-ray data, including 3D reconstruction and unsupervised anomaly detection. I hold a dual B.S. in Biomedical Engineering and Artificial Intelligence from Korea University, and currently also research at OGQ.
DL
