MaureenZOU / TSAM

[CVPR2021] Official Implementation for Progressive Temporal Feature Alignment Network for Video Inpainting
90 stars 11 forks source link

Object Mask - experimental settings. #6

Closed donglao closed 3 years ago

donglao commented 3 years ago

Thanks for sharing your work!

When removing foreground objects using object masks on the DAVIS dataset, there are no ground truth results available. How is the inpainting accuracy (PSNR, SSIM, VFID) evaluated?

MaureenZOU commented 3 years ago

I am guessing the wording may be confused. Object mask does have ground truth (The original frames), while object removal does not have ground truth. Hope this solve your confusion : )