UNet-AF: An alias-free UNet for image restoration
This work addresses a specific issue in image restoration for researchers and practitioners using UNet-based models, but it is incremental as it builds on existing translation-equivariant layers.
The authors tackled the problem of aliasing in UNet architectures, which hinders translation equivariance in image restoration tasks, and proposed an alias-free UNet that achieves competitive performance with a significant increase in measured equivariance.
The simplicity and effectiveness of the UNet architecture makes it ubiquitous in image restoration, image segmentation, and diffusion models. They are often assumed to be equivariant to translations, yet they traditionally consist of layers that are known to be prone to aliasing, which hinders their equivariance in practice. To overcome this limitation, we propose a new alias-free UNet designed from a careful selection of state-of-the-art translation-equivariant layers. We evaluate the proposed equivariant architecture against non-equivariant baselines on image restoration tasks and observe competitive performance with a significant increase in measured equivariance. Through extensive ablation studies, we also demonstrate that each change is crucial for its empirical equivariance. Our implementation is available at https://github.com/jscanvic/UNet-AF