Projects
Interpretable AI
Ongoing
Developed interpretable-by-design image classification models based on large language models (LLMs) and vision-language models (VLMs), aiming to make model decisions transparent without sacrificing accuracy.
Physics-Inspired Structured Pruning
2023 – 2024
Developed Electrostatic Force and Gravity regularization methods for efficient neural networks — using physical analogies to guide which filters can be removed with minimal loss of information for faster deep learning models.
Medical Image Analysis with Deep Learning
2021 – 2025
Designed lightweight and improved deep learning architectures for the automated analysis of medical images, including pediatric wrist fracture detection in X-ray images, intracranial hemorrhage detection in CT images, and low-dose CT denoising.