XPixelGroup / HAT

CVPR2023 - Activating More Pixels in Image Super-Resolution Transformer Arxiv - HAT: Hybrid Attention Transformer for Image Restoration
Apache License 2.0
1.14k stars 134 forks source link

difference between HAT_SRx4_ImageNet-pretrain.pth and HAT_SRx4.pth #99

Open maopian2333 opened 9 months ago

maopian2333 commented 9 months ago

hi, could i ask the difference between HAT_SRx4_ImageNet-pretrain.pth and HAT_SRx4.pth? I couldn't find an introduction for different model versions. Thanks a lot.

chxy95 commented 9 months ago

@maopian2333 Using or not using ImageNet pretraining. Mentioned in the paper.