Dear author, what an amazing work on unconditional image inpainting. But I have a concern and suggestion: Brushnet-like architecture has shown to achieve superior result compared to traditional sd-inpainting architecture, so it would be great to see your method combined with Brushnet just like PowerPaint-v2.
For that, it would be great if you can open-source the training code so we can try on new architecture
Dear author, what an amazing work on unconditional image inpainting. But I have a concern and suggestion: Brushnet-like architecture has shown to achieve superior result compared to traditional sd-inpainting architecture, so it would be great to see your method combined with Brushnet just like PowerPaint-v2.
For that, it would be great if you can open-source the training code so we can try on new architecture