search

yuval-alaluf / stylegan3-editing

Official Implementation of "Third Time's the Charm? Image and Video Editing with StyleGAN3" (AIM ECCVW 2022) https://arxiv.org/abs/2201.13433
https://yuval-alaluf.github.io/stylegan3-editing/
MIT License
660 stars 72 forks source link

Whether this project can be applied to other data sets like shoes #21

Closed ww451575464 closed 2 years ago

ww451575464 commented 2 years ago

hello, My team collected a lot of pictures of shoes，and we want to know if this project can help us generate more shoes。 thanks you ！

yuval-alaluf commented 2 years ago

It seems like what you want is to original StyleGAN3 repository here: https://github.com/NVlabs/stylegan3 You can use that repo to train a generator to generate new images of shoes. If you would like, you can then use this repo to edit those images, for example.

ww451575464 commented 2 years ago

thank you ! But my picture is unaligned and have no landmarks, I want to use this project to edit the generated picture

yuval-alaluf commented 2 years ago

You don't need aligned data or landmarks to train the StyleGAN3 generator. Once you train the generator on your data, you can then use this repo for editing.