김윤호 연구실

3D volume reconstruction and characterization of structural flexibilities from cryogenic electron microscopy. Focuses on determining conformational changes in flexible particles from 2D noisy images.

MR (Magnetic Resonance) imaging restoration including denoising and removing non-uniformity artifacts. Non-invasive imaging method using RF pulses for human body visualization without tissue damage.

Image restoration, decomposition and segmentation using PDE and variational frameworks for texture-noise separation and image segmentation applications.

Development and analysis of unified variational frameworks for solving eigenvalue/vector problems and phase retrieval problems in computational mathematics.

Reservoir Computing Networks as an alternative to Deep Learning. RCN leverages complex dynamical systems in the learning process with advantages over traditional deep learning approaches.

Mass conserving Allen-Cahn equations and their analytical and numerical properties. Study of phase field models for various applications in computational mathematics.