MUMU: Bootstrapping Multimodal Image Generation from Text-to-Image Data Paper • 2406.18790 • Published Jun 26 • 33 • 3