An Effective and Efficient Face Mask Recognition System for Edge Devices
Xuan Li · 2022 5th International Conference on Data Science and Information Technology (DSIT) · 2022
Wearing face masks is a simple but effective measure to mitigate the spread of COVID-19. Instead of manually searching persons without a face mask, designing an intelligent system to automatically detect and recognize them is necessary. The system is usually deployed on portable edge devices in order to be used in complex environments. As these devices have limited memory and computing power, popular large models are not suitable. In this paper, we propose a novel face mask recognition system, which consists of four modules: face detection, mask recognition, liveness detection, and face recognition. Experimental results verify that the proposed system has low latency, low memory burden, and satisfactory performance.