14.Common skin disease classification and recognition based on multi-model fusion

Abstract:

This study introduces a hybrid model classification approach for identifying common skin diseases, as a single convolutional neural network model may not ensure the complete accuracy of skin microscopy image recognition results and has weak generalization ability. Initially, data augmentation was employed to counterbalance the dataset

Keywords: Dermoscopy images;Multi-model fusion;Image classification and recognition;convolutional neural networks