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

AmazUtah_NLP at SemEval-2024 Task 9: A MultiChoice Question Answering System for Commonsense Defying Reasoning


This paper has been withdrawn by Qiwei Zhu

View a PDF of the paper titled GDSR: Global-Detail Integration through Dual-Branch Network with Wavelet Losses for Remote Sensing Image Super-Resolution, by Qiwei Zhu and 5 other authors

No PDF available, click to view other formats

Abstract: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 reconstruction 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 Wavelet Loss, a loss function that effectively captures high-frequency detail information in images, thereby enhancing the visual quality of SR, particularly in terms of detail reconstruction. Extensive experiments on several benchmarks, including AID, AID_CDM, RSSRD-QH, and RSSRD-QH_CDM, demonstrate that GSDR outperforms the state-of-the-art Transformer-based method HAT by an average of 0.05 dB in PSNR, while using only 63% of its parameters and 51% of its FLOPs, achieving an inference speed 2.9 times faster. Furthermore, the Wavelet Loss shows excellent generalization across various architectures, providing a novel perspective for RSI-SR enhancement.

Submission history

From: Qiwei Zhu [view email]
[v1]
Tue, 31 Dec 2024 10:43:19 UTC (2,865 KB)
[v2]
Tue, 7 Jan 2025 14:19:35 UTC (1 KB) (withdrawn)



Source link
lol

By stp2y

Leave a Reply

Your email address will not be published. Required fields are marked *

No widgets found. Go to Widget page and add the widget in Offcanvas Sidebar Widget Area.