MVDiffusion: How to take a pre-trained text2image model for a perspective view (e.g., Stable…
Yasutaka Furukawa
Visit on X@YasutakaFuruka12.5K followers
33,484 views
MVDiffusion: How to take a pre-trained text2image model for a perspective view (e.g., Stable Diffusion) and retrain to generate multiple consistent views (e.g., a panorama). Project site: Hugging Face demo: Code out in a month.
8 Comments
Comments from the original post on X (Twitter)

@YasutakaFuruka1
I am very sorry. Typo fixing. Had to delete the old one and retweet.

@YasutakaFuruka1
I had to delete and repost. Highly appreciate it if you could reshare/retweet this post
1 like

@raywzy1
Awesome work!!!!! I wonder where we could find the paper😊
1 like

@YasutakaFuruka1
Thank you Ziyu. Did not realize that we haven't uploaded to arxiv yet... Asking students to upload in a day.
1 like

@AIDailyGuy
I would add to my newsletter
1 like

@ptsi
🔥

@asrlhhh
does it support image prompt as input or only text prompt?

@YasutakaFuruka1
Only text for this work. But it seems trivial to add image-prompt capability.