Table 6 Ablation study results on road condition dataset.
From: An intelligent YOLO and CNN-BiGRU framework for road infrastructure based anomaly assessment
Model variant | mAP@0.5 (%) | AUROC (%) | F1-score | Latency (ms) |
|---|---|---|---|---|
Baseline (MLP only) | 78.12 | 0.802 | 0.763 | 15.2 |
YOLO + MLP | 91.54 | 0.889 | 0.857 | 11.8 |
YOLO + CNN-BiGRU | 94.67 | 0.917 | 0.893 | 9.8 |
DT + YOLO + CNN-BiGRU | 96.92 | 0.942 | 0.925 | 9.5 |
