Add SSIM as Loss Function

[bluehyena]

🚀 The feature, motivation and pitch

Hello, I am doing research in the field of Unsupervised Anomaly Detection. In Unsupervised Anomaly Detection, you can segment an anomaly area through Reconstruction Error, but in the case of areas with ambiguous boundaries, better performance is shown when using loss using SSIM than using existing losses. Therefore, I would like to be able to use SSIM as the loss function of pytorch. Thank you :)

Alternatives

No response

Additional context

No response

[janeyx99]

Thanks for raising this up! PyTorch Core has some pretty strict rules when it comes to adding a new Optimizer or Module function or feature: https://github.com/pytorch/pytorch/wiki/Developer-FAQ#i-have-a-new-function-or-feature-id-like-to-add-to-pytorch-should-it-be-in-pytorch-core-or-a-library-like-torchvision.

Currently, SSIM seems best suited to be implemented in a separate library (or using an existing one), and for the future, we'll happily reconsider if this function increases in popularity. We can leave this issue open to gauge popularity among other researchers and users.