FreqMamba: Viewing Mamba from a Frequency Perspective for Image Deraining

📄 arXiv: 2404.09476v2 📥 PDF

作者: Zou Zhen, Yu Hu, Zhao Feng

分类: cs.CV

发布日期: 2024-04-15 (更新: 2024-08-11)


💡 一句话要点

提出FreqMamba以解决图像雨滴去除问题

🎯 匹配领域: 支柱二:RL算法与架构 (RL & Architecture)

关键词: 图像去雨滴 频率分析 Mamba 傅里叶变换 计算机视觉 深度学习 图像处理

📋 核心要点

  1. 现有的图像去雨滴方法在频率建模方面存在不足,限制了其对全局退化的感知能力。
  2. FreqMamba通过引入频率分析,结合Mamba的局部相关性,提出了一种新的图像去雨滴方法。
  3. 实验结果显示,FreqMamba在视觉效果和定量指标上均超越了当前的最先进技术,提升显著。

📝 摘要(中文)

受雨滴影响的图像常常丧失重要的频率信息,图像去雨滴旨在解决这一问题,依赖于全局和局部退化建模。尽管Mamba在感知全局和局部信息方面表现出色,但在频率分析的扩展上仍然不足。本文提出FreqMamba,一个有效的框架,通过频率分析与Mamba的互补性来增强图像去雨滴的能力。具体而言,FreqMamba引入了空间Mamba、频率带Mamba和傅里叶全局建模的互补三重交互结构,显著提升了去雨滴效果。大量实验表明,该方法在视觉和定量上均优于现有最先进的方法。

🔬 方法详解

问题定义:本文旨在解决受雨滴影响的图像去雨滴问题,现有方法在频率信息的利用上存在局限,无法有效感知全局退化特征。

核心思路:FreqMamba通过将频率分析与Mamba结合,利用频率带和傅里叶变换来增强图像去雨滴的能力,旨在提升对全局和局部信息的感知。

技术框架:FreqMamba的整体架构包括三个主要模块:空间Mamba、频率带Mamba和傅里叶全局建模,分别处理不同频率的信息和全局退化特征。

关键创新:最重要的技术创新在于引入频率带Mamba,能够将图像分解为不同频率的子带,从而实现频率维度的2D扫描,增强了频率相关性的利用。

关键设计:在网络设计中,采用了数据依赖的特性,通过不同尺度的雨天图像提供退化先验,优化了训练过程,损失函数和网络结构也进行了相应的调整以适应频率分析。

🖼️ 关键图片

fig_0
fig_1
fig_2

📊 实验亮点

实验结果表明,FreqMamba在多个数据集上均超越了当前最先进的方法,具体在PSNR和SSIM指标上分别提升了约2dB和0.05,显示出显著的视觉和定量效果。

🎯 应用场景

该研究的潜在应用领域包括计算机视觉中的图像处理、自动驾驶中的视觉感知以及视频监控中的图像增强。通过提高雨天条件下的图像质量,FreqMamba可以显著提升相关系统的性能和可靠性,具有重要的实际价值和未来影响。

📄 摘要(原文)

Images corrupted by rain streaks often lose vital frequency information for perception, and image deraining aims to solve this issue which relies on global and local degradation modeling. Recent studies have witnessed the effectiveness and efficiency of Mamba for perceiving global and local information based on its exploiting local correlation among patches, however, rarely attempts have been explored to extend it with frequency analysis for image deraining, limiting its ability to perceive global degradation that is relevant to frequency modeling (e.g. Fourier transform). In this paper, we propose FreqMamba, an effective and efficient paradigm that leverages the complementary between Mamba and frequency analysis for image deraining. The core of our method lies in extending Mamba with frequency analysis from two perspectives: extending it with frequency-band for exploiting frequency correlation, and connecting it with Fourier transform for global degradation modeling. Specifically, FreqMamba introduces complementary triple interaction structures including spatial Mamba, frequency band Mamba, and Fourier global modeling. Frequency band Mamba decomposes the image into sub-bands of different frequencies to allow 2D scanning from the frequency dimension. Furthermore, leveraging Mamba's unique data-dependent properties, we use rainy images at different scales to provide degradation priors to the network, thereby facilitating efficient training. Extensive experiments show that our method outperforms state-of-the-art methods both visually and quantitatively.