What question did this study set out to answer?

To enhance semantic segmentation in remote sensing images by addressing challenges in scale variations and spatial distributions.

August 7, 2026Remote SensingOpen Access

SPFMamba: A Mamba-Based Network with Semantic Prompt and Frequency-Adaptive Fusion for Remote Sensing Image Semantic Segmentation

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial demonstrates improved segmentation accuracy in remote sensing images, indicating advanced modeling capabilities.

To enhance semantic segmentation in remote sensing images by addressing challenges in scale variations and spatial distributions.
Developed SPFMamba architecture combining semantic prompting and frequency-adaptive fusion.
Introduced semantic prompt global–local Mamba block for enhanced feature reconstruction.
Implemented high-frequency adaptive fusion module to refine responses in shallow layers.
SPFMamba achieves favorable segmentation accuracy on ISPRS Vaihingen, ISPRS Potsdam, and OpenEarthMap datasets.
Demonstrates improved preservation of object structures and fine spatial details.
Employs only 17.85 M parameters for efficient performance.

Wang et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark