MobileNet-Pascal-VOC2012

parkinson-mps

This project predicts motor score changes in Parkinson’s Disease patients using LFP signals pre- and post-DBS. Features like beta power difference were extracted to train models including Random Forest and XGBoost, achieving over 92% accuracy. Data and model scripts, along with analysis, are included.

multi-spectral-imaging

Multispectral Imaging with CAVE Dataset: code and data for my latest project exploring multispectral imaging technology using the CAVE dataset. Includes implementations of spectral unmixing, NMF, and sparse representation algorithms. Use for research or applications at your own risk.