aMUSEd is a lightweight masked image model (MIM) for text-to-image generation that is faster and more interpretable than latent diffusion. It can be fine-tuned to learn additional styles and is effective for large-scale text-to-image generation.

https://arxiv.org/abs//2401.01808

YouTube: https://www.youtube.com/@ArxivPapers

TikTok: https://www.tiktok.com/@arxiv_papers

Apple Podcasts: https://podcasts.apple.com/us/podcast/arxiv-papers/id1692476016

Spotify: https://podcasters.spotify.com/pod/show/arxiv-papers