This project develops an automated, deep learning-based behavioral analysis system to identify Alzheimer's disease (AD) features. By eliminating human bias and enhancing precision, it accelerates early screening and drug development for neurodegenerative diseases.

-

Using universal action space technology, our system extracts motion features without wearable sensors or pose estimation. By analyzing behaviors directly through big data, it seamlessly adapts to applications like home care, electronic geofencing, anomaly detection, and smart farming.