Skin Lesion Classification Using Densely Connected Convolutional Network

Classification of dermatoscopic image has become an engrossing research topic due to the necessity of earlier identification of particular diseases. Deep learning is performing a significant role to find a more efficient way of dermoscopic analysis. In this paper, DenseNet-121 was used to classify 7 types of skin lesions using the “HAM10000” (Human Against Machine) dataset. Data augmentation was used to make performance of the classifier more efficient. This research can have a positive impact to support dermatologist in the clinic to make more accurate decision in case of skin lesion identification.