Table 1.

The main parameters of the SSD network model.

Layer	Convolutional Kernel Size	Convolutional Kernel Number	Step Size	Filling	Feature Map Size
Conv1_1	3 × 3	64	1	1	300 × 300
Conv1_2	3 × 3	64	1	1	300 × 300
Maxpool1	2 × 2	1	2	0	150 × 150
Conv2_1	3 × 3	128	1	1	150 × 150
Conv2_2	3 × 3	128	1	1	150 × 150
Maxpool2	2 × 2	1	2	0	75 × 75
Conv3_1	3 × 3	256	1	1	75 × 75
Conv3_2	3 × 3	256	1	1	75 × 75
Conv3_3	3 × 3	256	1	1	75 × 75
Maxpool3	2 × 2	1	2	0	38 × 38
Conv4_1	3 × 3	512	1	1	38 × 38
Conv4_2	3 × 3	512	1	1	38 × 38
Conv4_3	3 × 3	512	1	1	38 × 38
Maxpool4	2 × 2	1	2	0	19 × 19
Conv5_1	3 × 3	512	1	1	19 × 19
Conv5_2	3 × 3	512	1	1	19 × 19
Conv5_3	3 × 3	512	1	1	19 × 19
Maxpool5	3 × 3	1	1	1	19 × 19
Conv6	3 × 3	1024	1	1	19 × 19
Conv7	1 × 1	1024	1	0	19 × 19
Conv8_1	1 × 1	256	1	0	19 × 19
Conv8_2	3 × 3	512	2	1	10 × 10
Conv9_1	1 × 1	128	1	0	10 × 10
Conv9_2	3 × 3	256	2	1	5 × 5
Conv10_1	1 × 1	128	1	0	5 × 5
Conv10_2	3 × 3	256	1	0	3 × 3
Conv11_1	1 × 1	128	1	0	3 × 3
Conv11_2	3 × 3	256	1	0	1 × 1