This document reviews various techniques for skin disease image recognition using deep learning. It summarizes 15 research papers that propose different methods for skin lesion segmentation, feature extraction, and classification to detect skin cancers like melanoma. Common techniques discussed include convolutional neural networks (CNNs), transfer learning models like VGG-16 and ResNet-50, and classifiers like support vector machines (SVMs). The proposed methodology will classify 5 types of skin cancers using the EfficientNet architecture for feature extraction from images and fully connected layers for classification. This aims to improve melanoma detection accuracy compared to other deep learning methods.