paint-by-inpaint/demo · Apply for community grant: Academic project (gpu)

paint-by-inpaint org Jun 28

•

This demo showcases our "Paint by Inpaint" approach for adding objects to images, trained with our PIPE dataset. This dataset is based on the insight that removing images is simpler than adding them. Using PIPE, we train a text following, editing diffusion model to add objects into images. Access to a GPU would significantly enhance the inference speed of our demo, allowing the community to experience our model firsthand and potentially propel further advancements in the field.

More details on our methods can be found here.
Paper page:
https://rotsteinnoam.github.io/Paint-by-Inpaint/
HuggingFace paper page:
https://huggingface.co/papers/2404.18212

hysts

Jun 29

Hi @noamrot , we've assigned ZeroGPU to this Space. Please check the compatibility and usage sections of this page so your Space can run on ZeroGPU.

noamrot

paint-by-inpaint org Jun 29

thanks!