Table 4.

Average validation recall, precision, $F1$-score, and AP@.5 for five trained Faster R-CNN and YOLOv5 models with color and motion-enhanced images.

Model	Dataset	Recall	Prec.	F1-Score	AP@.5
FR-CNN	Color	0.867	0.889	0.878	0.890
FR-CNN	Motion	0.889	0.862	0.875	0.900
YOLOv5	Color	0.888	0.897	0.892	0.914
YOLOv5	Motion	0.919	0.852	0.884	0.924