Reproduce “FSG-Net: Frequency-Spatial Synergistic Gated Network for High-Resolution Remote Sensing Change Detection” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Zhongxiang Xie, Shuangxi Miao, Yuhan Jiang, Zhewei Zhang, Jing Yao, Xuecao Li, Jianxi Huang, Pedram Ghamisi
A frequency-spatial synergistic gated network, FSG-Net, was developed to advance high-resolution remote sensing change detection. It achieves state-of-the-art performance by effectively suppressing pseudo-changes and delineating sharp change boundaries across challenging public datasets, maintaining a balanced computational footprint.
Key insights
- Synergistic processing in both frequency and spatial domains is crucial for simultaneously mitigating pseudo-changes and enhancing the precision of genuine change detection.
- Adaptive interaction strategies tailored for different frequency sub-bands (LL, LH, HL, HH) effectively disentangle semantic changes from task-irrelevant variations.
- A semantics-driven gating mechanism enables selective fusion of multi-level features, allowing for accurate propagation of fine-grained spatial details to achieve sharp change boundaries.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.