What happens if you do too much 'layers.MaxPooling2D()'

Question

So Im making this CNN and I was wondering what happens if you do too much layers.MaxPooling2D() because I searched up what pooling does and it reduces the spatial size of the images aka downsamples it, so if I do too much pooling I could end up with a photo 1 pixel big.

So should I be careful with layers.MaxPooling2D()?

Answer 1

Basically max pooling reduces dimentions. Absolutely, you should construct your model carefully and also care size and stride parameters of max pooling operation.

Here is the good explanations of basic operations:

https://developers.google.com/machine-learning/practica/image-classification/convolutional-neural-networks?hl=id