Could you provide the image pairs you have used for reproducing the result

[fangxuehouwuming]

Could you provide the image pairs you have used for reproducing the result

[fahadshamshad]

We use the image pairs provided in the link below:

CGCL-codes/AMT-GAN#15

For CelebA, we report the average results on 1,000 images (as provided by AMT-GAN). Similarly, for LADN, we report the results after dividing the LADN images into four groups, as described in the paper.