VQCNIR: Clearer Night Image Restoration with Vector-Quantized Codebook

TL;DR

VQCNIR introduces a novel data-driven approach for night image restoration using a vector-quantized codebook, achieving consistent and high-quality results by integrating adaptive illumination and cross-attention modules.

Contribution

The paper proposes VQCNIR, a new night image restoration method that leverages a vector-quantized codebook and novel modules for better illumination and detail preservation.

Findings

Achieves state-of-the-art performance on synthetic and real datasets.

Effectively restores details and illumination in low-light images.

Demonstrates robustness and consistency across various benchmarks.

Abstract

Night photography often struggles with challenges like low light and blurring, stemming from dark environments and prolonged exposures. Current methods either disregard priors and directly fitting end-to-end networks, leading to inconsistent illumination, or rely on unreliable handcrafted priors to constrain the network, thereby bringing the greater error to the final result. We believe in the strength of data-driven high-quality priors and strive to offer a reliable and consistent prior, circumventing the restrictions of manual priors. In this paper, we propose Clearer Night Image Restoration with Vector-Quantized Codebook (VQCNIR) to achieve remarkable and consistent restoration outcomes on real-world and synthetic benchmarks. To ensure the faithful restoration of details and illumination, we propose the incorporation of two essential modules: the Adaptive Illumination Enhancement Module (AIEM) and the Deformable Bi-directional Cross-Attention (DBCA) module. The AIEM leverages the inter-channel correlation of features to dynamically maintain illumination consistency between degraded features and high-quality codebook features. Meanwhile, the DBCA module effectively integrates texture and structural information through bi-directional cross-attention and deformable convolution, resulting in enhanced fine-grained detail and structural fidelity across parallel decoders. Extensive experiments validate the remarkable benefits of VQCNIR in enhancing image quality under low-light conditions, showcasing its state-of-the-art performance on both synthetic and real-world datasets. The code is available at https://github.com/AlexZou14/VQCNIR.