In CNN, the convolution and pooling layers may change the size of the feature map. A simple CNN is given below with input image size of 32x32x3. In each layer, several filters are applied with different filter size (f), stride (s), and padding (p). Compute the missing values (a, b, c, d, e, f). a= b= POOL1 Max Pool d filters OTÖTÖTT? f=2 s=2 14x14xc s=1 axaxb p=0 C= d= e= 32 x 32 x 3 6 filters f=5 s=1 CONV1 p=0 f=5 CONV2 p=0 Max Pool f=2 POOL2 0= 10x10x16 5=2 exexf p=0

In CNN, the convolution and pooling layers may change the size of the feature map. A simple CNN is given below with input image size of 32x32x3. In each layer, several filters are applied with different filter size (f), stride (s), and padding (p). Compute the missing values (a, b, c, d, e, f). a= b= POOL1 Max Pool d filters OTÖTÖTT? f=2 s=2 14x14xc s=1 axaxb p=0 C= d= e= 32 x 32 x 3 6 filters f=5 s=1 CONV1 p=0 f=5 CONV2 p=0 Max Pool f=2 POOL2 0= 10x10x16 5=2 exexf p=0