Official implementation code accompanying the paper: "Learning Multimodal Affinities for Textual Editing in Images". - View it on GitHub