Skip to main content
. 2018 Aug 29;9(9):4509–4526. doi: 10.1364/BOE.9.004509

Fig. 3.

Fig. 3

Network architecture for CNN-S with number of columns in B-scan = 400 and 3 target surfaces. N=number of kernels used in the convolution layer, CONV=Convolution Layer, FC=Fully connected layer, n=number of neurons in the FC layer.