Params confusion

How the param is 0 for first pooling layer ?
According to him as calculated in other layers, the params should be (f * f * depthOfInput * K) ie. (2 * 2 * 6) * 6 = 144. How that is 0 ?

The parameters are 0 in CNN architecture in max-pooling layer (not the convolutional layer), because in max-pooling layer there are no trainable parameters (just maximum value from the patch is passed on to next layer).

Thank you, he told later in the lecture .

1 Like