Tensorflow.js tf.pool() Function

Introduction: Tensorflow.js is an open-source library that is developed by Google for running machine learning models as well as deep learning neural networks in the browser or node environment.

The .pool() function is used to execute an N-D pooling functioning.

Syntax:

tf.pool(input, windowShape, poolingType, pad, dilations?, strides?)

Parameters:

• input: The stated input tensor which is either of rank 4 or else rank 3 and of shape: [batch, height, width, inChannels]. Moreover, in case the rank is 3, then the batch of size 1 is presumed. It can be of type tf.Tensor3D, tf.Tensor4D, TypedArray, or Array.
• windowShape: The stated filter size: [filterHeight, filterWidth]. It can be of type [number, number], or number. In case, filterSize is a singular number, then filterHeight == filterWidth.
• poolingType: The stated type of pooling, it can be either 'max' or 'avg'.
• pad: The stated type of algorithm for padding. It can be of type valid, same, number, or conv_util.ExplicitPadding.
  1. Here, for same and stride 1, the output would have identical size as input, irrespective of the filter size.
  2. For, 'valid' the output shall be smaller than the input in case, the filter size is larger than 1*1x1.
• dilations: The stated dilation rates: [dilationHeight, dilationWidth] in that the input values are sampled over the height as well as width dimensions in dilated pooling. The by default value is [1, 1]. Moreover, in case dilations is a single number, then dilationHeight == dilationWidth. And if its greater than 1, then all the values of the strides should be 1. It is optional and is of type [number, number], number.
• strides: The stated strides of the pooling: [strideHeight, strideWidth]. In case, strides is a singular number, then strideHeight == strideWidth. It is optional and is of type [number, number], or number.

Return Value: It returns tf.Tensor3D or tf.Tensor4D.

Example 1:

// Importing the tensorflow.js library
import * as tf from "@tensorflow/tfjs"

// Defining input tensor
const x = tf.tensor3d([1, 2, 3, 4], [2, 2, 1]);

// Calling pool() method
const result = tf.pool(x, 3, 'avg', 'same', [1, 2], 1);
 
// Printing output
result.print();

Output:

Tensor
    [[[0.4444444],
      [0.6666667]],

     [[0.4444444],
      [0.6666667]]]

Example 2:

// Importing the tensorflow.js library
import * as tf from "@tensorflow/tfjs"

// Calling pool() method
tf.tensor3d([1.2, 2.1, 3.0, -4], [2, 2, 1]).pool(3,
                    'conv_util.ExplicitPadding', 1, 1).print();
 

Output:

Tensor
    [[[3],
      [3]],

     [[3],
      [3]]]

Reference: https://js.tensorflow.org/api/latest/#pool