A Simple Framework for Contrastive Learning of Visual Representations paper: https://arxiv.org/abs/2002.05709