search

Dixin-Lab / generalized-face-landmarker

Official PyTorch implementation for the paper Generalizable Face Landmarking Guided by Conditional Face Warping (CVPR 2024).
11 stars 1 forks source link

Why not using torch.grid_sample in dense_image_warp function? #1

Open nttstar opened 2 months ago

nttstar commented 2 months ago

Thanks

plustwo0 commented 2 months ago

We employ the warping following WarpGAN. It's an alternative way to warp the image using torch.grid_sample in the Line 60 of warp.py instead of dense_image_warp() function, as long as the warping process is differentiable.

Thanks for the advice!