Mask detection based on R2-yolov5

Properly wearing masks in specific situations can effectively inhibit virus transmission, while manual detection can be labor-intensive and inefficient. To solve this problem, a target detection algorithm based on R2-yolov5 is proposed. Based on yolov5, the residual module bottleneck in the unique C3 structure is replaced by the Res2Net residual module and optimized in terms of channels and activation functions, and CBAM module is introduced to improve the detection effect. Through experiments, the map of R2-yolov5 proposed in this paper can reach 88.9% tested on the mask dataset, which is 2.4% better than the yolov5 network, and there is a significant improvement in the detection effect.