Announcement_12

We show that learning discrete representations of human movements leads to better sensor-based Human Activity Recognition (HAR) in Towards Learning Discrete Representations via Self-Supervision for Wearables-Based Human Activity Recognition.