HAT: Hybrid Attention Transformer for Image Restoration

TL;DR

HAT is a novel hybrid attention transformer that combines channel and window-based self-attention to enhance input utilization and cross-window information aggregation, significantly improving performance across various image restoration tasks.

Contribution

The paper introduces a Hybrid Attention Transformer (HAT) that integrates channel and window-based self-attention with an overlapping cross-attention module and a pre-training strategy for superior image restoration.

Findings

Achieves state-of-the-art results on multiple benchmarks.

Effectively utilizes more input pixels for better restoration.

Extends to diverse applications like super-resolution and denoising.

Abstract

Transformer-based methods have shown impressive performance in image restoration tasks, such as image super-resolution and denoising. However, we find that these networks can only utilize a limited spatial range of input information through attribution analysis. This implies that the potential of Transformer is still not fully exploited in existing networks. In order to activate more input pixels for better restoration, we propose a new Hybrid Attention Transformer (HAT). It combines both channel attention and window-based self-attention schemes, thus making use of their complementary advantages. Moreover, to better aggregate the cross-window information, we introduce an overlapping cross-attention module to enhance the interaction between neighboring window features. In the training stage, we additionally adopt a same-task pre-training strategy to further exploit the potential of the model for further improvement. Extensive experiments have demonstrated the effectiveness of the proposed modules. We further scale up the model to show that the performance of the SR task can be greatly improved. Besides, we extend HAT to more image restoration applications, including real-world image super-resolution, Gaussian image denoising and image compression artifacts reduction. Experiments on benchmark and real-world datasets demonstrate that our HAT achieves state-of-the-art performance both quantitatively and qualitatively. Codes and models are publicly available at https://github.com/XPixelGroup/HAT.