ELEGANT icon indicating copy to clipboard operation
ELEGANT copied to clipboard

Can ELEGANT perform inter-class translation?

Open imlixinyang opened this issue 6 years ago • 6 comments

imlixinyang avatar Dec 26 '19 03:12 imlixinyang

Another question is that it is shown that ELEGANT can perform multi-label translation, but I wonder whether it is performed by two-step translation or a single step (exchange two codes in or not in the meantime)?

imlixinyang avatar Dec 26 '19 04:12 imlixinyang

I guess the question that you are asking is: can we swap the hairstyle of two images with bangs? I did not try that, but I think ELEGANT can do that.

It is done in a single step translation. We claim that ELEGANT can control multiple attributes simultaneously. See Fig. 9.

Prinsphield avatar Dec 26 '19 05:12 Prinsphield

Thank you for your reply, I'll have a try later. It would be amazing if it works because these two tasks are not involved in the training.

imlixinyang avatar Dec 26 '19 06:12 imlixinyang

I guess the question that you are asking is: can we swap the hairstyle of two images with bangs? I did not try that, but I think ELEGANT can do that.

It is done in a single step translation. We claim that ELEGANT can control multiple attributes simultaneously. See Fig. 9.

hello @Prinsphield About the second question, it means exchanging two attribute codes of two reference images with the code of input image in the meantime? I wonder how is this achieved? because the model is performed on a reference image and an input image.

Jessicall avatar Jan 17 '20 07:01 Jessicall

I guess the question that you are asking is: can we swap the hairstyle of two images with bangs? I did not try that, but I think ELEGANT can do that. It is done in a single step translation. We claim that ELEGANT can control multiple attributes simultaneously. See Fig. 9.

hello @Prinsphield About the second question, it means exchanging two attribute codes of two reference images with the code of input image in the meantime? I wonder how is this achieved? because the model is performed on a reference image and an input image.

Hello @Jessicall, since different attribute codes are disentangled represented in the latent encodings, so we can exchange the encodings of the input and two reference images simultaneously. You may refer to the code for details.

Prinsphield avatar Jan 17 '20 08:01 Prinsphield

Thanks for your reply!

Jessicall avatar Jan 17 '20 08:01 Jessicall