CAPformer: Compression-Aware Pre-trained Transformer for Low-Light Image Enhancement

TL;DR

CAPformer is a novel transformer-based model designed for low-light image enhancement that explicitly accounts for JPEG compression effects, improving image quality in resource-limited settings.

Contribution

It introduces a compression-aware pre-training strategy and a brightness-guided self-attention mechanism to better handle compressed low-light images.

Findings

Outperforms existing LLIE methods under compression.

Effectively mitigates JPEG compression artifacts.

Enhances low-light images with preserved details.

Abstract

Low-Light Image Enhancement (LLIE) has advanced with the surge in phone photography demand, yet many existing methods neglect compression, a crucial concern for resource-constrained phone photography. Most LLIE methods overlook this, hindering their effectiveness. In this study, we investigate the effects of JPEG compression on low-light images and reveal substantial information loss caused by JPEG due to widespread low pixel values in dark areas. Hence, we propose the Compression-Aware Pre-trained Transformer (CAPformer), employing a novel pre-training strategy to learn lossless information from uncompressed low-light images. Additionally, the proposed Brightness-Guided Self-Attention (BGSA) mechanism enhances rational information gathering. Experiments demonstrate the superiority of our approach in mitigating compression effects on LLIE, showcasing its potential for improving LLIE in resource-constrained scenarios.