Hello, and thank you for sharing the code for your pix2pix model.
I tried generating images using the provided model, but the results I obtained were noticeably lower in quality compared to those shown in your paper. I wanted to kindly ask if you could provide some guidance on the following points:Is the released model checkpoint the same one used for the results shown in the paper?
Are there any additional settings or preprocessing steps required to reproduce the results more accurately?
Are there specific hyperparameters or conditions (input image size..) that significantly affect output quality?
Any help or clarification would be greatly appreciated.
Thank you again for making your work publicly available.
Best regards,
Jihyun Kim
Hello, and thank you for sharing the code for your pix2pix model.
I tried generating images using the provided model, but the results I obtained were noticeably lower in quality compared to those shown in your paper. I wanted to kindly ask if you could provide some guidance on the following points:Is the released model checkpoint the same one used for the results shown in the paper?
Are there any additional settings or preprocessing steps required to reproduce the results more accurately?
Are there specific hyperparameters or conditions (input image size..) that significantly affect output quality?
Any help or clarification would be greatly appreciated.
Thank you again for making your work publicly available.
Best regards,
Jihyun Kim