Mask detection algorithm based on fused C3-CBAM attention mechanism

The invention provides a mask detection algorithm based on a fused C3-CBAM (Channel Attention Module) attention mechanism on the basis of deep learning, a CBAM attention mechanism module fused with Channel Attention and Spatial Attention is introduced into a C3 module on the basis of a YOLOv5 target detector, the algorithm detection performance is improved while fewer parameters are introduced in the calculation process, the algorithm detection efficiency is improved, and the algorithm detection efficiency is improved. And the SIoU-NMS is adopted to optimize the redundant candidate prediction frame, so that effective detection of the mouth mask wearing behavior of the crowd is realized, and the monitoring system can automatically analyze the crowd condition, thereby assisting the monitoring personnel in maintaining public security. 本发明以深度学习为基础提出一种基于融合C3-CBAM（Convolutional Block Attention Module）注意力机制的口罩检测算法，以YOLOv5目标检测器为基础，在C3模块中引入了通道注意力（Channel Attention）与空间注意力（Spatial Attention）融合的CBAM注意力机制模块，在计算过程中引入较少参数的同