adobe-research / sam_inversion

[CVPR 2022] GAN inversion and editing with spatially-adaptive multiple latent layers
Other
170 stars 10 forks source link

Add Replicate demo and API #8

Closed ariel415el closed 2 years ago

ariel415el commented 2 years ago

Hey @GaParmar and @junyanz ! 👋

Good work on this project I was very impressed by the reconstruction quality of the suggested inversion process. The images actually look better that the original in some cases !

This pull request creates an online cloud demo and API for this repository on Replicate.com. These are based on a docker enviroment built for this repository and makes it easier for other people to run it.

You can view the demo here: https://replicate.com/adobe-research/sam_inversion And running the demos functionatlity with our API is as simple as:

import replicate
model = replicate.models.get("[adobe-research/sam_inversion]")
model.predict(image_path=, image_category=, edit_direction=, change_norm=, thresh=)

By clicking on "Claim this model" on the demo site you'll be able to edit the demo site, add/remove examples of your liking and we'll feature it on our website and tweet about it too.

In case you're wondering who I am, I'm from Replicate, where we're trying to make machine learning reproducible. We got frustrated that we couldn't run all the really interesting ML work being done. So, we're going round implementing models we like. 😊

cjchapman commented 2 years ago

@ariel415el I'm not involved in the development of this code, so I defer to @junyanz and @GaParmar on this PR.