Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Regarding the evaluation of reconstruction quality #1

Open
Sierkinhane opened this issue Oct 16, 2024 · 1 comment
Open

Regarding the evaluation of reconstruction quality #1

Sierkinhane opened this issue Oct 16, 2024 · 1 comment

Comments

@Sierkinhane
Copy link

Hi, thank you for the nice work. :)

I'd like to know whether you have evaluated the reconstruction quality on the commonly used benchmarks like ImageNet and UCF. And how about the reconstruction quality of the small faces and texts.

@wilson1yan
Copy link
Contributor

We evaluate our model on ImageNet in the paper, but not on UCF. UCF in the code is primarily there as an example if you want to train your own model.

If I remember correctly, the pretrained models can do smaller faces / texts to some extent, but still struggle if they're too small. The pretrained models are also available to use if you want to try on your own data of interest

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants