HAR using wearable sensors

A large number of vendors have wearable watches and bracelets with fitness trackers. These watches and bracelets have GPS, accelerometer, gyroscope, heart rate sensor, and/or ambient light sensors. Employing sensor fusion, they combine the output of these sensors to make a prediction about the activity. Due to the temporal nature of the data, it is a challenging time series classification task.

Fitbit (https://www.fitbit.com/smarttrack), a premier company in the field of fitness trackers, use a technology it calls SmartTrack, which recognizes activities with continuous movement or light movement. It uses the intensity and patterns of the movement to classify the activity. It classifies the activity in seven classes, ...

Get Hands-On Artificial Intelligence for IoT now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.