Heterogeneous window transformer for image denoising
This work addresses computational efficiency in image denoising for applications requiring fast processing, though it is incremental as it builds on existing transformer-based methods.
The paper tackles image denoising by proposing a heterogeneous window transformer (HWformer) to balance distance modeling and computational efficiency, achieving a 30% reduction in denoising time compared to Restormer.
Deep networks can usually depend on extracting more structural information to improve denoising results. However, they may ignore correlation between pixels from an image to pursue better denoising performance. Window transformer can use long- and short-distance modeling to interact pixels to address mentioned problem. To make a tradeoff between distance modeling and denoising time, we propose a heterogeneous window transformer (HWformer) for image denoising. HWformer first designs heterogeneous global windows to capture global context information for improving denoising effects. To build a bridge between long and short-distance modeling, global windows are horizontally and vertically shifted to facilitate diversified information without increasing denoising time. To prevent the information loss phenomenon of independent patches, sparse idea is guided a feed-forward network to extract local information of neighboring patches. The proposed HWformer only takes 30% of popular Restormer in terms of denoising time.