查看文章

Skin diseases are one of the most common diseases in the world but still its accurate diagnosis is exceptionally challenging because of its complexities of skin tone, color, lesion size, presence of hair. This paper provides an approach to use various Deep Learning based computer vision based techniques to automatically predict the various kinds of skin diseases. The system uses a novel custom Convolutional Neural Network (CNN) model disease which successfully predicts the skin disease if it belongs to 7 classes of diseases the model has been trained on. The system consists of three phases-The feature extraction phase, the training phase and the testing/validation phase. The model makes use of deep learning technology to train itself with the various skin images from the HAM10000 open source dataset. The main objective of this model is to achieve maximum accuracy of skin disease prediction which it successfully achieves a training accuracy of 94.81% and test accuracy of 80.52%. This model is further deployed as open source software using a Tensorflow. js.