Phillip Isola

Isola co-created pix2pix, which made image-to-image translation a widely used generative technique, and helped establish contrastive learning for visual representations. At MIT he now studies shared representations, embodied intelligence and how general capabilities emerge from interaction.