WEARING DETECTION OF MEDICAL PROTECTIVE EQUIPMENT BASED ON IMPROVED YOLOV7

This study proposes an improved algorithm model based on YOLOv7 to further improve the detection accuracy of wearing medical protective equipment correctly. In this work, the CBAM attention mechanism was added to the YOLOv7 backbone network with the channel and spatial attention mechanism modules combined to improve its ability to pay attention on the medical protective equipment. In addition, EIoU was utilized as the loss function to speed up the regression speed of the prediction box and improve the robustness of the model. The experimental results show that compared with original YOLOv7, the accuracy and mAP@0.5 of this improved method on the self-made medical protective equipment wearing dataset are increased by 6.8 and 5.2 percentage points, respectively. Therefore, the improved algorithm model based on YOLOv7 proposed in this paper can effectively improve the detection efficiency of wearing medical protective equipment correctly.