EF-SwinNet: A Hybrid EfficientNet-Swin Transformer Model for Skin Cancer Classification

Skin cancer ranks as the most common type of cancer worldwide. In this study, we present EF -SwinNet, a hybrid architecture combining EfficientNet and swin transformer models, designed to classify skin lesions with high accuracy. The dataset used for training and testing the model is HAMI0000, which is known for its imbalance. In order to tackle this problem, we implemented several data augmentation methods to successfully alleviate the problem of classification imbalance. Our empirical findings illustrate the exceptional efficacy of the suggested hybrid model, attaining an average accuracy of 98 % and an Fl score of 96%. Furthermore, we employed the Grad-CAM technique to offer a deeper understanding of the learning process of the model by graphically representing the significance of various features. Finally, we analyze the performance of our model by comparing it with current cutting-edge methods, emphasizing its improvements in the classification of skin cancer.