You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I note that the quantitative results in your new NTED paper is inconsistent with GFLA,can you explain this and point out the difference in your metrics calculation.
NTED
GFLA
The text was updated successfully, but these errors were encountered:
Hi,
Thanks for your interest. This difference is caused by the evaluation images. In this paper, we use the resized High-Resolution images in DeepFashion as our evaluation images.
The FID is calculated by comparing the distribution of the generated images with the distribution of the images in the training set.
The SSIM and LPIPS are calculated between the generated images and the corresponding GT images.
We will provide the evaluation scripts as soon as possible.
Thanks for the excellent work @RenYurui. It'd be really nice if you can share the evaluation scripts. Also, if I understood correctly the Table.1 is for the pose control. Did you also calculate the scores for appearance control?
I note that the quantitative results in your new NTED paper is inconsistent with GFLA,can you explain this and point out the difference in your metrics calculation.
NTED
GFLA
The text was updated successfully, but these errors were encountered: