An Efficient License Plate Detection Approach Using Lightweight Deep Convolutional Neural Networks

Table 2

The structure of ESPNetv2 used in this paper.

Layer	Type	Output size
Layer 1 (L1)	Standard convolution (3 × 3 × 32, stride 2)	160 × 160
Layer 2 (L2)	1 × strided EESP	80 × 80
Layer 3 (L3)	1 × strided EESP	40 × 40
	3 × EESP
Layer 4 (L4)	1 × strided EESP	20 × 20
	7 × EESP
Layer 5 (L5)	1 × strided EESP	10 × 10
	3 × EESP