HAHR Reference Architecture

Overview

A reference architecture for Human Activity and Health Recognition (HAHR), a principled scaffold that spans the full sensing-to-inference pipeline across four distinct sensing modalities.

Approach

The architecture covers four sensing lineages: wearable inertial (IMU-based activity and gait), bio-impedance (body composition and physiological state), ambient RF sensing (WiFi CSI-based presence and motion), and vision-based recognition. It is designed to support building HAHR systems from macro behavioral signals down to micro physiological indicators.