[IMPC] Special Session: Biological Application

[IMPC-04] Real-Time Fall Detection for Embedded System Using Deep Learning

*Chih-Chieh Yang¹, Chieh-En Lee¹, Tzu-Yuan Huang¹, Chung-Hao Tien¹ (1. National Yang Ming Chiao Tung Univ. (Taiwan))

[Presentation Style] Online

In this study, we developed an embedded visual system which can real-time detect the falls by depth video in multiple views. Based on MoViNets, we successfully achieved 99.21% accuracy under 380 testing events.