Research

Our research focuses on the intersection of computational imaging, signal processing, and machine learning.

1) Cryo-Electron Microscopy Reconstruction

We study computational phase retrieval and reconstruction under low-dose imaging conditions, aiming to improve robustness and enable higher-resolution structural biology.

Topics

• computational phase retrieval for cryo-EM micrographs and movies
• modeling CTF effects and zero-crossings
• noise-aware reconstruction and uncertainty quantification

2) Computational Microscopy

We develop physical forward models and reconstruction algorithms for lens-free and phase imaging, targeting wide FOV, high resolution, and efficient computation.

Topics

• diffraction-based imaging and propagation modeling
• ptychography and near-field imaging
• adaptive sampling and high-NA computational imaging

3) Physics-Informed Deep Learning for Inverse Problems

We build learning systems that respect imaging physics and remain reliable across conditions.

Topics

• optimization-inspired networks (unrolled / plug-and-play)
• controllable / parameterized models for varying imaging settings
• generalization under distribution shift

4) Imaging Under Low-Dose / Photon-Starved Conditions

We investigate modeling and algorithms for extreme-noise regimes.

Topics

• Poisson-Gaussian noise modeling
• dose weighting and motion effects
• robust training objectives and evaluation protocols