Skin Disease Classification Using Mobilenet-RseSK Network

Abstract The traditional deep learning method has a large amount of calculation, long time training and complex network structure, which is not easy to be applied embedded or mobile devices. To solve the these problems, we proposed an improved lightweight named as Mobilenet-RseSK for skin disease classification. Firstly, new attention mechanism seSK module proposed, used replace position SE in original network. This can better perform feature extraction improve performance than module. Secondly, using RBN normalization, maintains advantages BN, strengthens representation specific features, degree accuracy identification. We compare with MobilenetV3, Ghost other advanced networks on HAM10000 dataset. promotes classification by 1.7% compared Compared mobilenet network, our achieves 85% test set. certain practical value classification, effective algorithm.