Consider the following input channels$(5$x$5$x$3),$ two$(3$x$3$x$3)$ filter volumes and two$(1$x$1)$bias$.$ A convolutional layer cross$-$correlates the input and kernels and adds a scalar bias toproduce an output. Assume that a zero padding$($P$=1),$ stride of $2$ is applied to the input volume.Illustrate the convolution layer computation and show the output$(3$x$3$x$2).[10$M$]$x$1$ x$2$ x$3111221202110122110221212110122110221212112122110221212111122111221222120122$w$00$ w$01$ w$021111-1-1-1-1-11-11111-10-11-11-111-1-1-1$w$10$ w$11$ w$1211-11-1-1-1-1-111-1-11-1-11-11-1-10-1-1-1-1-1$b$0$ b$11-1$