luoxuan-cs / pama Goto Github PK

Hi @luoxuan-cs,
I am interested in your research. So I have some questions:

1. Why content loss is different from style loss. Why don't use self-similarity loss for style loss?
2. In Eq. 8, can I use mean and std instead of mean and covariance?

Hope to hear from you soon.

python main.py eval
the content and sample images are the example on website https://huggingface.co/spaces/akhaliq/PAMA
and the weights is download from google drive named <original_PAMA.zip>
The program does not report an error
but I got a gray image

And I found that the website provided could not return results, and each time it ended with error

The performance of VGG19 is nice, but I want to use a lighter encoder, How can I retrain an encoder (for example mobilenet v3)?
Thank you~

please add a license file