利用区域信息 Transformer 进行单图像去雨
Exploiting Regional Information Transformer for Single Image Deraining
AI总结：
针对单图像去雨中雨区与非雨区混合处理导致特征混淆的问题，提出 Regformer，通过区域掩码注意力和混合门控前馈模块实现分区建模与局部补偿，达到最先进去雨性能。
摘要：
基于 Transformer 的单图像去雨（Single Image Deraining，SID）方法已取得显著成功，这主要归功于其捕捉长程交互的强大能力。然而，我们注意到，当前方法同时处理受雨影响区域和未受雨影响区域，忽视了这些区域之间的差异，导致雨条纹与背景部分混淆，并且无法获得有效的交互，最终造成次优的去雨结果。为解决上述问题，我们提出 Region Transformer（Regformer），这是一种新颖的 SID 方法，强调独立处理受雨影响和未受雨影响区域的重要性，同时考虑二者的联合影响以实现高质量图像重建。该方法的核心是创新的 Region Transformer Block（RTB），它集成了 Region Masked Attention（RMA）机制和 Mixed Gate Forward Block（MGFB）。我们的 RTB 用于受雨影响与未受雨影响区域的注意力选择，以及混合尺度的局部建模。RMA 生成针对这两类区域及其交互定制的注意力图，使模型能够捕捉去雨所需的全面特征。为更好地恢复高频纹理并捕捉更多局部细节，我们开发 MGFB 作为补偿模块，以完成局部混合尺度建模。大量实验表明，我们的模型达到最先进性能，显著提升了图像去雨质量。我们的代码和训练模型均已公开发布。
英文摘要：
Transformer-based Single Image Deraining (SID) methods have achieved remarkable success, primarily attributed to their robust capability in capturing long-range interactions. However, we've noticed that current methods handle rain-affected and unaffected regions concurrently, overlooking the disparities between these areas, resulting in confusion between rain streaks and background parts, and inabilities to obtain effective interactions, ultimately resulting in suboptimal deraining outcomes. To address the above issue, we introduce the Region Transformer (Regformer), a novel SID method that underlines the importance of independently processing rain-affected and unaffected regions while considering their combined impact for high-quality image reconstruction. The crux of our method is the innovative Region Transformer Block (RTB), which integrates a Region Masked Attention (RMA) mechanism and a Mixed Gate Forward Block (MGFB). Our RTB is used for attention selection of rain-affected and unaffected regions and local modeling of mixed scales. The RMA generates attention maps tailored to these two regions and their interactions, enabling our model to capture comprehensive features essential for rain removal. To better recover high-frequency textures and capture more local details, we develop the MGFB as a compensation module to complete local mixed scale modeling. Extensive experiments demonstrate that our model reaches state-of-the-art performance, significantly improving the image deraining quality. Our code and trained models are publicly available.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
