Skip to main content

Table 3 Inference latency comparison with 4K resolution image on GPU and Jetson Xavier

From: Automatically search an optimal face detector for a specific deployment environment

 

Latency (ms)

Avg. latency (ms)

Xavier

2080Ti

Xavier

2080Ti

TinaFace [49]

1363.40

155.62

1362.15

153.45

EXTD-32 [19]

417.24

51.50

417.24

51.50

LFFD [20]

104.41

14.57

98.63

11.84

Yolov5s [25]

182.80

18.84

182.58

18.27

Retina0.25 [4]

52.54

6.85

52.21

5.34

Our-small

26.81

2.93

23.91

1.70

Our-large

34.36

3.61

31.12

2.26