Using combined Soft-NMS algorithm Method with Faster R-CNN model for Skin Lesion Detection

The detection of skin diseases has always been a hot topic in the medical field. With the development of deep learning, more and more neural network models have been used in medical research and have achieved good results. In this paper, based on the existing target detection model Faster R-CNN, we replace the NMS algorithm in it with Soft-NMS. The experimental results verify the effectiveness of our improvement. Compared with Faster R-CNN, our method can frame the skin disease area more accurately by reducing the misrecognized area of non-lesion areas. At the same time, our method can better deal with the situation of blurred boundaries of skin diseases. The data set we used comes from ISIC (International Skin Imaging Collaboration).