Embedded Monitoring System for Preventing Lonely Death Based on Edge AI

In this paper, we propose an monitoring system employing an edge AI module. The proposed system reduces latency and alleviates data leakage issue with an edge AI module. In addition, power efficiency was optimized by implementing a wake-up function depending on human detection results. In order to verify the feasibility of the entire system, we implemented the edge AI module on field programmable gate array (FPGA). The accuracy of movement detection was measure at 0.953.