Research Article
A New Multiface Target Detection Algorithm for Students in Class Based on Bayesian Optimized YOLOv3 Model
Table 2
Comparison of calculation results of different methods.
| Algorithm | IoU = 0.5 missed detection rate | mAP (%) IoU = 0.5 | AP (%) | Detection speed (Fps) | Parameter (MB) | IoU = 0.6 | IoU = 0.7 | IoU = 0.8 |
| Fast R-CNN | 15.8 | 84.1 | 81.2 | 79.9 | 76.8 | 45.5 | 15.32 | FaceNet | 11.3 | 88.6 | 86.3 | 83.8 | 81.2 | 36.6 | 18.61 | DeepID | 10.8 | 89.2 | 86.5 | 83.9 | 82.1 | 38.3 | 17.85 | SSD | 12.2 | 87.8 | 85.3 | 82.9 | 80.4 | 40.5 | 14.83 | YOLOv1 | 9.8 | 90.1 | 88.2 | 86.5 | 83.9 | 41.3 | 15.65 | YOLOv2 | 9.5 | 90.5 | 89.3 | 87.5 | 85.2 | 43.7 | 16.21 | YOLOv3 | 8.9 | 91.0 | 90.1 | 89.5 | 88.7 | 46.2 | 17.52 | Ours | 7.8 | 92.2 | 91.3 | 90.9 | 90.1 | 45.1 | 18.35 |
|
|