PulseTrendingJournal ClubResearchersJournalsExplore
Instagram
HomeTrendingJournal ClubExplore
Synapse
⌘+K
Synapse
February 2, 2026Remote SensingOpen Access

Lite-BSSNet: A Lightweight Blueprint-Guided Visual State Space Network for Remote Sensing Imagery Segmentation

View Full Paper
Ask AI
Bookmark
Share

Authors

JYJiaxin YanYXYuxiang XieTLTao Li

Discussion

Loading...

Member takes

Overview

Experiments demonstrate Lite-BSSNet improves segmentation accuracy in remote sensing imagery, suggesting enhanced model efficiency.

Key Points

  • The research aims to enhance remote sensing image segmentation by addressing global context and local detail challenges.
  • Developed a Structural Blueprint Generator for edge-enhanced structural blueprints.
  • Introduced a Visual State Space Bridge for aligning multi-level features and enhancing edge signals.
  • Implemented a Structural Repair Block to improve the receptive field and reduce upsampling artifacts.
  • Lite-BSSNet achieved the highest segmentation accuracy among lightweight models with mIoU of 83.9% on ISPRS Vaihingen and 86.7% on Potsdam.
  • Required only 45.4 GFLOPs, indicating a favorable accuracy-efficiency trade-off.

Cite This Study

Yan et al. (2026) studied this question.

synapsesocial.com/papers/6980ff19c1c9540dea811d2chttps://doi.org/10.3390/rs18030441
View Full Paper
Ask AI
Bookmark
Share