DDT: Dual-branch Deformable Transformer for Image Denoising

TL;DR

The paper introduces DDT, a dual-branch deformable transformer that efficiently captures local and global features for image denoising, achieving state-of-the-art results with reduced computational costs.

Contribution

It proposes a novel dual-branch deformable transformer architecture that models local and global interactions in parallel for improved image denoising.

Findings

Achieves state-of-the-art denoising performance.

Reduces computational complexity compared to existing transformer methods.

Effective on both real-world and synthetic noise datasets.

Abstract

Transformer is beneficial for image denoising tasks since it can model long-range dependencies to overcome the limitations presented by inductive convolutional biases. However, directly applying the transformer structure to remove noise is challenging because its complexity grows quadratically with the spatial resolution. In this paper, we propose an efficient Dual-branch Deformable Transformer (DDT) denoising network which captures both local and global interactions in parallel. We divide features with a fixed patch size and a fixed number of patches in local and global branches, respectively. In addition, we apply deformable attention operation in both branches, which helps the network focus on more important regions and further reduces computational complexity. We conduct extensive experiments on real-world and synthetic denoising tasks, and the proposed DDT achieves state-of-the-art performance with significantly fewer computational costs.