Utility for own model require a retrained Imagenet Pretrained ?

[JarveeLee]

Thanks for your great work.
Models applied in other task usually require a Imagenet Pretrained backbones.
If I want to use this Module into own backbone , is it necessary to train a ImageNet Pretrain ... ?
Or just replace MaxPool with MaxBlurPool then load the original Pretrain weight, then to train other task ?
For example you don't have Res101 in weights/download_antialiased_models.sh , but I hope to farely compare Res101 with the same Pretrained weights.

[richzhang]

Yes -- models should be retrained (or fine-tuned) after adding the antialiasing layers. I just added antialiased ResNet101 models.

[JarveeLee]

Thanks~

[richzhang]

I added instructions for loading an antialiased model (simple 2 lines of code!). I think this covers it, so I'm closing this issue now. Thanks for your interest in the project!

[dongzhi0312]

Thanks for your great work.
Models applied in other task usually require a Imagenet Pretrained backbones.
If I want to use this Module into own backbone , is it necessary to train a ImageNet Pretrain ... ?
Or just replace MaxPool with MaxBlurPool then load the original Pretrain weight, then to train other task ?
For example you don't have Res101 in weights/download_antialiased_models.sh , but I hope to farely compare Res101 with the same Pretrained weights.

can you share the weights of author offered? I cannot download, thanks

[richzhang]

They should be listed in this file. Thanks!