The following code demonstrates max pooling on a tensor using a VALID padding scheme:
import tensorflow as tfbatch_size=1input_height = 3input_width = 3input_channels = 1def main(): sess = tf.InteractiveSession() layer_input = tf.constant([ [ [[1.0], [0.2], [2.0]], [[0.1], [1.2], [1.4]], [[1.1], [0.4], [0.4]] ] ])# The strides will look at the entire input by using the image_height and image_widthkernel = [batch_size, input_height, input_width, input_channels]max_pool = tf.nn.max_pool(layer_input, kernel, [1, 1, 1, 1], "VALID")print(sess.run(max_pool))if __name__ == '__main__': main()
The output of the preceding listing will give the maximum values in the window 3 x 3 x 1:
[[[[ 2.]]]]
The following diagram explains how max pool ...
