DL4J machine learning model to predict chances that skin defects are melanoma, seborrheic keratosis, or nevus (moles). Model used is YOLOv2 modified to predict the aforementioned 3 output classes. Trained and evaluated model set to access laptop webcam to test real-time identification
Challenge link: https://www.kaggle.com/c/siim-isic-melanoma-classification/overview
Dataset link: https://www.kaggle.com/yihfoo/labelled-balanced-skin-lesion-dataset