Open Access
Table 4
Comparison of CIA-YOLO with methods in related fields.
| Type | Method | Function | Accuracy |
|---|---|---|---|
| Physics-based method [4] | AE technique + signal analysis | Detecting bridge cable breaks | The error is approximately 5%. |
| ML-based method [5] | ECT signals + Machine Learning | classify defect type and shape | Over 90% |
| DL-based method [11] | Color Segmentation + R-CNN | Adapts to various defect types | 90.61% |
| DL-based method [21] | K-means++ + ECA + Focal Loss | Detecting minor damage and complex structures | 92.2% |
| DL-based method [22] | YOLOv5s + C3CBAM + C3Ghost | Detect and classify corrosion on metal surfaces | 95.6% |
| DL-based method [38] | YOLOv8 + CAFM + DyConv + EfficientViT | Detecting defects in brake joint welds | 90.5% |
| Proposed method | CIA-YOLO | Detecting and classifying wire rope defects | 95.2% |
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.