SIR-Former: Stereo Image Restoration Using Transformer

Zizheng Yang,Mingde Yao,Jie Huang,Man Zhou,Feng Zhao

DOI: https://doi.org/10.1145/3503161.3548177

2022-01-01

Abstract:Stereo image pairs record the scene from two different views and introduce cross-view information for image restoration. However, there are two challenges in utilizing the cross-view information for stereo image restoration: cross-view alignment and information fusion. Most existing methods adopt convolutional neural networks to align the views and fuse the information locally, which has difficulty in capturing the global correspondence across stereo images for view alignment and makes it hard to integrate the long-term information across views. In this paper, we propose to address the stereo image restoration with transformer by leveraging its powerful capability of modeling long-range context dependencies. Specifically, we construct a stereo image restoration transformer (SIR-Former) to effectively exploit the cross-view correlations. First, to explore the global correspondence for view alignment effectively, we devise a stereo alignment transformer (SAT) module across stereo images, enabling robust alignment under the epipolar constraint. Then, we design a stereo fusion transformer (SFT) module for aggregating the cross-view information in a small horizontal neighborhood, aiming to enhance important features for succeeding restoration. Extensive experiments show that SIR-Former can remarkably boost quantitative and qualitative quality on various image restoration tasks (e.g., super-resolution, deblurring, deraining, and low-light enhancement), which demonstrate the effectiveness of the proposed framework.