Explain what the following code does di_erently from the almost identical code at the beginning of Section 3.4.3: convOut = tf.nn.conv2d(image, flts, [1,1,1,1], "SAME") convOut =...

Explain what the following code does di_erently from the almost identical code at the beginning of Section 3.4.3:

convOut = tf.nn.conv2d(image, flts, [1,1,1,1], "SAME")

convOut = tf.nn.maxpool(convOut, [1,2,2,1], [1,1,1,1], "SAME").

In particular, for an arbitrary values of image and flts, does convOut have same shape, in both cases? Does it necessarily have the same values? Is one set of values a proper subset of the other? In each case, why or why not?