December 5, 2025Applied SciencesOpen Access

EMS-UKAN: An Efficient KAN-Based Segmentation Network for Water Leakage Detection of Subway Tunnel Linings

View Full Paper

Ask AI

Bookmark

Authors

Discussion

Proposed deep learning model improves detection of water leakage in subway tunnel linings, suggesting enhanced efficiency and performance.

Achieving competitive segmentation performance, the model effectively reduces computational complexity for real-world applications.
The Efficient Multi-Scale U-shaped KAN-based Segmentation Network addresses limitations of traditional deep learning methods.
Key innovations include depthwise separable convolutions and enhanced feature extraction to capture multi-scale features.
Extensive experiments on the Tunnel Water Leakage dataset validate the model's robustness in detecting water leakage.

Tan et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark

Proposed deep learning model improves detection of water leakage in subway tunnel linings, suggesting enhanced efficiency and performance.