标题： GDSR：通过小波损失的双分支网络进行全局细节集成的遥感图像超分辨率 显示英文标题

标题： GDSR: Global-Detail Integration through Dual-Branch Network with Wavelet Losses for Remote Sensing Image Super-Resolution

摘要： 近年来，深度神经网络，包括卷积神经网络、Transformer 和状态空间模型，在遥感图像（RSI）超分辨率（SR）方面取得了显著进展。然而，现有的 SR 方法通常忽略了全局和局部依赖之间的互补关系。这些方法要么专注于捕捉局部信息，要么优先考虑全局信息，这导致模型无法同时有效地捕捉全局和局部特征。此外，当应用于大规模 RSI 时，它们的计算成本变得难以承受。为了解决这些挑战，我们引入了接收权重键值（RWKV）在 RSI-SR 中的新应用，该方法以线性复杂度捕捉长距离依赖关系。为了同时建模全局和局部特征，我们提出了全局-细节双分支结构 GDSR，通过并行 RWKV 和卷积操作来处理大规模 RSI 实现 SR。此外，我们在两个分支之间引入了全局-细节重建模块（GDRM），以弥补它们的互补作用。此外，我们提出了双组多尺度小波损失，通过双组子带策略和跨分辨率频率对齐的小波域约束机制，以提高 RSI-SR 的重建保真度。在 AID、UCMerced 和 RSSRD-QH 等多个基准数据集上，使用两种退化方法进行的大量实验表明，GSDR 在 PSNR 上平均比最先进的基于 Transformer 的方法 HAT 提高了 0.09 dB，同时仅使用其 63% 的参数和 51% 的 FLOPs，推理速度提高了 3.2 倍。 显示英文摘要

摘要： In recent years, deep neural networks, including Convolutional Neural Networks, Transformers, and State Space Models, have achieved significant progress in Remote Sensing Image (RSI) Super-Resolution (SR). However, existing SR methods typically overlook the complementary relationship between global and local dependencies. These methods either focus on capturing local information or prioritize global information, which results in models that are unable to effectively capture both global and local features simultaneously. Moreover, their computational cost becomes prohibitive when applied to large-scale RSIs. To address these challenges, we introduce the novel application of Receptance Weighted Key Value (RWKV) to RSI-SR, which captures long-range dependencies with linear complexity. To simultaneously model global and local features, we propose the Global-Detail dual-branch structure, GDSR, which performs SR by paralleling RWKV and convolutional operations to handle large-scale RSIs. Furthermore, we introduce the Global-Detail Reconstruction Module (GDRM) as an intermediary between the two branches to bridge their complementary roles. In addition, we propose the Dual-Group Multi-Scale Wavelet Loss, a wavelet-domain constraint mechanism via dual-group subband strategy and cross-resolution frequency alignment for enhanced reconstruction fidelity in RSI-SR. Extensive experiments under two degradation methods on several benchmarks, including AID, UCMerced, and RSSRD-QH, demonstrate that GSDR outperforms the state-of-the-art Transformer-based method HAT by an average of 0.09 dB in PSNR, while using only 63% of its parameters and 51% of its FLOPs, achieving an inference speed 3.2 times faster.