Table 5 Metrics for each pathology type (YOLOv8l-seg)
From: Deep learning-driven pathology detection and analysis in historic masonry buildings of Suzhou
Class | Images | Instances | Box | Mask | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
Precision | Recall | mAP50 | mAP50-95 | Precision | Recall | mAP50 | mAP50-95 | |||
all | 174 | 1963 | 0.53 | 0.516 | 0.486 | 0.274 | 0.504 | 0.484 | 0.459 | 0.25 |
ML | 3 | 23 | 0.765 | 0.706 | 0.756 | 0.425 | 0.764 | 0.696 | 0.756 | 0.472 |
DD | 148 | 1001 | 0.406 | 0.536 | 0.431 | 0.288 | 0.403 | 0.521 | 0.416 | 0.261 |
CR | 59 | 144 | 0.57 | 0.479 | 0.465 | 0.232 | 0.555 | 0.458 | 0.425 | 0.17 |
SS | 137 | 746 | 0.499 | 0.489 | 0.485 | 0.319 | 0.498 | 0.479 | 0.47 | 0.287 |
BI | 20 | 49 | 0.408 | 0.367 | 0.291 | 0.105 | 0.301 | 0.265 | 0.228 | 0.061 |