LayeredSense: Hierarchical Recognition of Complex Daily Activities Using Wearable Sensors
Chak Man Lam
TL;DR
LayeredSense addresses the challenge of recognizing complex daily activities by decomposing signals into unit patterns and inferring higher-level activities from their distributions. Using data from a Myo armband's IMU streams, the method employs sliding-window feature extraction, Gaussian Mixture Models for new pattern discovery, and Random Forest classifiers for unit-pattern and activity recognition. It introduces a semi-supervised workflow for discovering new unit patterns with user labeling to expand the training set, and a bag-of-words approach to aggregate unit-pattern sequences into activity predictions. The results show high offline accuracy for unit-pattern recognition (about 98% with cross-validation) and robust online performance (roughly 82–83%) for both unit patterns and activities, indicating scalability for real-time daily activity monitoring.
Abstract
Daily activity recognition has gained prominence due to its applications in context-aware computing. Current methods primarily rely on supervised learning for detecting simple, repetitive activities. This paper introduces LayeredSense, a novel framework designed to recognize complex activities by decomposing them into smaller, easily identifiable unit patterns. Utilizing a Myo armband for data collection, our system processes inertial measurement unit (IMU) data to identify basic actions like walking, running, and jumping. These actions are then aggregated to infer more intricate activities such as playing sports or working. LayeredSense employs Gaussian Mixture Models for new pattern detection and machine learning algorithms, including Random Forests, for real-time activity recognition. Our system demonstrates high accuracy in identifying both unit patterns and complex activities, providing a scalable solution for comprehensive daily activity monitoring