Research Article
Attention-Based Convolutional Neural Network for Pavement Crack Detection
Table 8
Yantai Dataset–Dice loss.
| | Model | Precision/(%) | Mean IoU/(%) | Recall/(%) | F1/(%) |
| | ENet | 94.80 | 82.17 | 92.10 | 93.43 | | ExFuse | 92.10 | 74.12 | 87.66 | 89.87 | | FCN | 90.23 | 77.65 | 89.16 | 89.69 | | LinkNet | 94.45 | 80.76 | 91.62 | 93.01 | | SegNet | 91.80 | 74.86 | 90.23 | 91.01 | | UNet | 93.76 | 80.11 | 91.10 | 92.41 |
| | CrackResAttentionNet | 95.43 | 82.75 | 94.2 | 94.81 |
|
|