Database

[OrangeZz0331]

Can you share your sketch database for us to train models？I want to try it.Thank you very much!

[puckikk1202]

The dataset is too large for direct upload. If you wish to test our results, you can download our pre-trained models from this link:google drive Alternatively, you can download the CelebA dataset directly from their website celeba, and extract the edge map from it to recreate the dataset.

[OrangeZz0331]

OK，Thank you very much！ 

橙橙子 @.***

 

------------------ 原始邮件 ------------------ 发件人: @.***>; 发送时间: 2023年12月6日(星期三) 中午12:01 收件人: "puckikk1202/difffacesketch2023"; 抄送: "Author"; 主题: Re: [puckikk1202/difffacesketch2023] Database (Issue #4)

The dataset is too large for direct upload. If you wish to test our results, you can download our pre-trained models from this link:google drive Alternatively, you can download the CelebA dataset directly from their website celeba, and extract the edge map from it to recreate the dataset.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

[OrangeZz0331]

And I would like to know if there is a section in the project code that converts a facial image with background cleared into a sketch image? If so, where is it stored? How can I obtain it if there is no need?

[puckikk1202]

Sorry for the inconvenience, we have not provided the code for cleaning the background. Maybe you can try the one with the background first if you only want to test the training. Alternatively, CelebAMask-HQ has provided the background masks that you can use for cleaning them.

[OrangeZz0331]

OK,Thank you very much!