PHAR – Practical Human Activity Recognition

PHAR explores human activity recognition using wearable and ambient sensors. It emphasizes practical deployment, robustness, and multi-modal sensor fusion.

Recognize daily human activities with low-cost sensors. Build robust classification models for real-world environments. Evaluate multimodal sensing approaches for activity inference.

Integrates motion, ambient, and context sensors. Emphasizes practical, deployable system design. Extends TECO’s long-standing interest in context-awareness.

Influences wearable health and behavior recognition systems. Provides early frameworks for ML-based activity detection. Basis for later rPPG and physiological monitoring research at TECO.