Skip to main content
. 2021 May 8;21(9):3263. doi: 10.3390/s21093263

Table 9.

The comparative experiments of AP of different models in three categories.

Sort Size IOU Face Face_Mask WMI
Proposed work 416 × 416 AP@.50 0.979 0.995 0.973
416 × 416 AP@.75 0.978 0.995 0.983
416 × 416 AP@.50:.95 0.767 0.939 0.834
YOLO-v4 416 × 416 AP@.50 0.943 0.969 0.944
416 × 416 AP@.75 0.680 0.899 0.800
416 × 416 AP@.50:.95 0.541 0.740 0.670
YOLO-v3 416 × 416 AP@.50 0.921 0.981 0.941
416 × 416 AP@.75 0.617 0.888 0.835
416 × 416 AP@.50:.95 0.559 0.789 0.724
SSD 300 × 300 AP@.50 0.941 0.986 0.988
300 × 300 AP@.75 0.503 0.920 0.926
300 × 300 AP@.50:.95 0.518 0.789 0.790
Faster R-CNN 600 × 600 AP@.50 0.943 0.974 0.950
600 × 600 AP@.75 0.700 0.927 0.866
600 × 600 AP@.50:.95 0.612 0.824 0.769