I would like to know how did we solve this question and obtained A1 and A2:

A neural network takes image f of size 55 as input and cutputs a clas fication label. The notwork passes appled to the twe generated actioction mape and the resuits are sent fo g tully corinected loper that outputs ane pameneter. Finaly, a sigmaid function is iaged to compute the output of the nehwork Y (a) Find the number of parameters in the network. Assume that the poels at the boundsy are ignored (b) Assume that pwote at the boundary are ignored when fitering, the 33 fitor weights are ghon by Sempute the festiting actoadion maps (a) Assume that non-corvolubon weglts are equal to 1, all biases are equal to 0 . Compute the output of the network. * is the convolution operation [a] number of parameters =(10+10)+(18+1)39 (b) size of the activation map =53+1=3 (c) After ReLU A2 will be a matrix of zero. x=i=118i=118wt+b=180,y=1+e1as1=1 A neural network takes image f of size 55 as input and cutputs a clas fication label. The notwork passes appled to the twe generated actioction mape and the resuits are sent fo g tully corinected loper that outputs ane pameneter. Finaly, a sigmaid function is iaged to compute the output of the nehwork Y (a) Find the number of parameters in the network. Assume that the poels at the boundsy are ignored (b) Assume that pwote at the boundary are ignored when fitering, the 33 fitor weights are ghon by Sempute the festiting actoadion maps (a) Assume that non-corvolubon weglts are equal to 1, all biases are equal to 0 . Compute the output of the network. * is the convolution operation [a] number of parameters =(10+10)+(18+1)39 (b) size of the activation map =53+1=3 (c) After ReLU A2 will be a matrix of zero. x=i=118i=118wt+b=180,y=1+e1as1=1