Table 6 Performance evaluation of representative object detection architectures in thermal imaging sensor sub-dataset.
From: MMFW-UAV dataset: multi-sensor and multi-view fixed-wing UAV dataset for air-to-air vision tasks
Methods | Recall | Precision | F1-Score | AP | FPS |
|---|---|---|---|---|---|
Yolo V4-Tiny | 97.19% | 99.36% | 0.98 | 99.08% | 413.95 |
Yolo V7-Tiny | 98.12% | 99.68% | 0.99 | 99.80% | 292.79 |
Faster-RCNN | 69.38% | 44.40% | 0.54 | 58.30% | 48.72 |
Yolo X | 97.50% | 99.68% | 0.99 | 98.91% | 245.20 |
Detr | 41.25% | 40.62% | 0.41 | 32.67% | 41.94 |
