Deep learning, the latest evolution of machine learning, has been promised to impact the world to the same degree as the internet. However, deploying an accurate deep-learning model on limited computational resources of embedded system is one of the state-of-the-art challenges. Gait Up is proud to introduce the first ever known deep-learning based human activity classifier using wearable technology.

Using a belt-worn Physilog 5, this deep-learning classifier achieves to detect different daily activities with an accuracy of more than 98%.
This edge computing application (computing close to where sensors gather data) improves reliability of data that is being processed and reduces latency of the application.
This super precise and low-power innovation powered by GaitUp opens new doors for long-term monitoring of daily-physical-activity.

Congrats to Foucauld Tabourin and Gait Up firmware team for the great job!