Fanghua-Yu
New version soon? Support for more samplers?
Will there be a newer, even more powerful version of SUPIR some time in 2025, and if so will it also support other/newer samplers?
Thank you
We are going to have a much more powerful model for both image and video announced in early 2025~
@JasonGUTU Wow that is very exciting to hear! Thank you for the prompt reply. When you say video, it will be able to enhance the individual frames and be temporally stable for video?
powerful model for both image and video
Will any of this be available in open source?
@JasonGUTU Hope progress is going well! Topaz recently released a cloud-based video diffusion upscale model called "Starlight", but I suspect your new version of SUPIR will be even better (and can hopefully be run locally)!
@JasonGUTU Hope all is well. Apologies if I am overeager, just wondering how progress is going on the new and improved model?
Teremos um modelo muito mais poderoso para imagem e vídeo anunciado no início de 2025~
We're past the middle of the year and still nothing...
Research and development cannot be rushed, and we spent a lot of time raising funds. But the good news is that our paper has been submitted.
our paper has been submitted
Oh, that's really cool. Can I get a link to the paper?
Research and development cannot be rushed, and we spent a lot of time raising funds. But the good news is that our paper has been submitted.
Very exciting news! @JasonGUTU Will the paper be publicly viewable at some point?