Adversarial Representation Learning for Text-to-Image Matching | IEEE Conference Publication | IEEE Xplore