Ting Chen

Chen led SimCLR, which showed that an image model could learn strong visual representations without human labels by comparing altered views of the same picture. The simple approach helped make self-supervised learning a standard part of computer vision.