August 21, 2025Open Access

HybridSeg: An Efficient Multi-Scale Mamba Architecture for Real-Time Semantic Segmentation in Railway Safety Monitoring Systems

View Full Paper

Ask AI

Bookmark

Achieving 84.8% mean Intersection over Union (mIoU) indicates successful semantic segmentation performance in railway safety monitoring.
The architecture integrates innovative techniques like Structure-Aware Preprocessing and Multi-Scale Feature Fusion, facilitating robust real-time performance.
Evaluation leveraged railway surveillance datasets, maintaining high pixel accuracy of 93.8% while operating at 31.8 frames per second.
The method supports key object classes, enhancing detection of personnel and infrastructure, crucial for safety improvements.

Discussion

No takes yet. Share an insight, caveat, or question.

HybridSeg demonstrates improved segmentation accuracy for railway safety in resource-constrained environments, suggesting better operational safety.

Fu et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark

HybridSeg demonstrates improved segmentation accuracy for railway safety in resource-constrained environments, suggesting better operational safety.