CLIP

Understanding and Mitigating Compositional Issues in Text-to-Image Generative Models