Semantic segmentation for crack detection via generative knowledge distillation DOI Creative Commons
Seungbo Shim

Automation in Construction, Journal Year: 2025, Volume and Issue: 175, P. 106201 - 106201

Published: April 24, 2025

Language: Английский

Dynamic context-aware high-resolution network for semi-supervised semantic segmentation DOI
Khawaja Iftekhar Rashid, Chenhui Yang, Chenxi Huang

et al.

Engineering Applications of Artificial Intelligence, Journal Year: 2025, Volume and Issue: 143, P. 110068 - 110068

Published: Jan. 15, 2025

Language: Английский

Citations

2
Topology-joint Curvilinear Segmentation Network using Confidence-based Bezier Topological Representation DOI
Jianwei Li, Yuchun Huang, Xi Ye

et al.

Engineering Applications of Artificial Intelligence, Journal Year: 2025, Volume and Issue: 143, P. 110045 - 110045

Published: Jan. 18, 2025

Language: Английский

Citations

0
A unified approach for weakly supervised crack detection via affine transformation and pseudo label refinement DOI Creative Commons

Zhongmin Huangfu,

Yibo Jiao,

Fupeng Wei

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: March 13, 2025

Consistent detection of cracks in engineering structures is essential for maintaining structural integrity. Deep neural networks perform well this discipline, although their pixel-level labeling reliance increases costs. Thus, weakly supervised learning methods have emerged. However, labels are substantially worse quality than those manual labeling. Current deep network visual interpretation approaches issues including erroneous target localization. This study proposes an Affine Transformation and Pseudo Label Refinement (AT-CAM) method. The methodology comprises three phases: the initial phase employs a geometric enhancement strategy to produce sequence enhanced images from input images, utilizing Axiom-based Grad-CAM (XGradCAM) algorithm generate class activation maps each image, which subsequently amalgamated into unified saliency map; subsequent phase, information flow pathways subsampling convolutional layer modified by designated Hook. samples utilized invert eliminate checkerboard noise produced during integration spatially; third stage, dynamic range compression mechanism employed augment prominence cracked areas compressing highlighted regions map diminishing influence background noise. experimental results indicate that method proposed segmentation accuracy 7.2% relative original baseline, markedly improves interpretability networks, offers novel, efficient, cost-effective, interpretable approach detecting engineering.

Language: Английский

Citations

0
Topology-informed deep learning for pavement crack detection: Preserving consistent crack structure and connectivity DOI

Jiayv Jing,

Ling Ding, Xu Yang

et al.

Automation in Construction, Journal Year: 2025, Volume and Issue: 174, P. 106120 - 106120

Published: March 21, 2025

Language: Английский

Citations

0
Semantic segmentation for crack detection via generative knowledge distillation DOI Creative Commons
Seungbo Shim

Automation in Construction, Journal Year: 2025, Volume and Issue: 175, P. 106201 - 106201

Published: April 24, 2025

Language: Английский

Citations

0