Table 4 Performance of YOLOv8-seg model variants (small, medium, large) on the same dataset
From: Deep learning-driven pathology detection and analysis in historic masonry buildings of Suzhou
Model | Box | Mask | ||||||
|---|---|---|---|---|---|---|---|---|
Precision | Recall | mAP@50 | mAP@50-95 | Precision | Recall | mAP@50 | mAP@50-95 | |
small | 0.566 | 0.446 | 0.462 | 0.25 | 0.62 | 0.395 | 0.437 | 0.231 |
medium | 0.563 | 0.475 | 0.452 | 0.26 | 0.549 | 0.463 | 0.437 | 0.234 |
large | 0.53 | 0.516 | 0.486 | 0.274 | 0.504 | 0.484 | 0.459 | 0.25 |
