Scholarlink

<p>In low-light scenarios, existing low-light image enhancement algorithms face significant challenges, such as over-enhancement of brightness, loss of details in high-light regions, and undesirable color distortions, which severely hinder their performance. To address these limitations, we propose an improved low-light image enhancement algorithm, Retinexformer on Performer Attention (RPA). By introducing an efficient attention mechanism to better handle illumination variations, RPA enhances the model’s capability to capture long-range global information and model local dependencies. The proposed framework comprises two key modules: a light decomposition module that generates illumination maps to brighten the image and a damage restoration module that leverages Relative-Performer Multi Self Attention (RPMSA) to refine image quality. Extensive experiments on six paired datasets and three unpaired datasets qualitatively and quantitatively demonstrate the effectiveness and generalization ability of RPA. Visually, it effectively suppresses high-light overexposure and preserves details and color consistency in high-contrast lighting conditions, resulting in more natural enhancement effects. This work provides a novel perspective to low-light image enhancement systems, improving their performance in real-world applications. Code is available at <a href="https://github.com/JJCcxk/RPA">https://github.com/JJCcxk/RPA</a>.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Enhancing low-light images with performer attention: a retinex-based approach

  • 期刊论文

  • 发表日期：

    2025年9月26日

查看全文
  • Yunxue Shao,
  • Yijin Diao,
  • Lingfeng Wang

摘要

In low-light scenarios, existing low-light image enhancement algorithms face significant challenges, such as over-enhancement of brightness, loss of details in high-light regions, and undesirable color distortions, which severely hinder their performance. To address these limitations, we propose an improved low-light image enhancement algorithm, Retinexformer on Performer Attention (RPA). By introducing an efficient attention mechanism to better handle illumination variations, RPA enhances the model’s capability to capture long-range global information and model local dependencies. The proposed framework comprises two key modules: a light decomposition module that generates illumination maps to brighten the image and a damage restoration module that leverages Relative-Performer Multi Self Attention (RPMSA) to refine image quality. Extensive experiments on six paired datasets and three unpaired datasets qualitatively and quantitatively demonstrate the effectiveness and generalization ability of RPA. Visually, it effectively suppresses high-light overexposure and preserves details and color consistency in high-contrast lighting conditions, resulting in more natural enhancement effects. This work provides a novel perspective to low-light image enhancement systems, improving their performance in real-world applications. Code is available at https://github.com/JJCcxk/RPA.