VICTR: Visual Information Captured Text Representation for Text-to-Image Multimodal Tasks

Open in new window