|
|
Heba successfully defended her PhD thesis:
This thesis introduces a new adaptive, hierarchical classification framework that significantly improves the accuracy, scalability, and interpretability of IMU‑based human activity recognition. By combining adaptive segmentation, multi‑mother wavelet feature extraction, and a global decision‑tree architecture, the approach handles complex real‑world movements with greater efficiency and lower computational cost. The work demonstrates strong performance across multiple datasets, paving the way for more reliable HAR systems in healthcare, aging support, and activity monitoring. Congratulations ! And a special thanks goes to Prof. Sreeraman Rajan who supervised this research for many years. |
