Tensorflow.js tf.layers.minimum() Function

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

The tf.layers.minimum() function is used to create a layer that is used to compute the element-wise minimum of an Array of inputs. It takes as input a list of tensors, having the same shape.

Syntax:

tf.layers.minimum (args)

Parameters:

It takes as input an object: args (Object). It is optional to provide args object as input. Following are the fields you can provide in the args object.

• inputShape ((null or number)[]): creates an input layer that is inserted before this layer.
• batchInputShape ((null or number)[]): has the same purpose as the above parameter but if both input shape and batchInputShape are defined, batchInputShape will be preferred.
• batchSize (number): if both the above parameters are not specified, batch Size is used to construct the batchInputShape.
• dtype : the data type of the layer. Eg: float32, int32, etc.
• name (string): it is used to give a name to the layer.
• weights (tf.Tensor[]): it provides initial weight values.
• trainable (boolean): it is used to specify whether the weights are updatable by fit. The default value is true.

Return Value: It returns element-wise minimum.

Example 1:

Javascript

const tf = require("@tensorflow/tfjs")

// providing input
const x = tf.input({shape: [4, 4, 4]});
const y = tf.input({shape: [4, 4, 4]});

// creating required layer
const minimumLayer = tf.layers.minimum();
const minimum = minimumLayer.apply([x, y]);
console.log(minimum.shape);

Output:

[ null, 4, 4, 4 ]

Example 2:

In this example, we will provide the args object as input with the fields of name and trainable.

Javascript

const tf = require("@tensorflow/tfjs")

// providing input
const x = tf.input({shape: [5, 5, 5]});
const y = tf.input({shape: [5, 5, 5]});
const z = tf.input({shape: [5, 5, 5]});

// creating required layer
const minimumLayer = tf.layers.minimum({name:"layer1", trainable:false});
const minimum = minimumLayer.apply([x, y, z]);
console.log(minimumLayer.name)
console.log(minimumLayer.trainable)
console.log(minimumLayer.shape);

Output:

layer1
false
[ null, 5, 5, 5 ]

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