Convolution operation concept in TensorFlow - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to create a 2D convolutional layer with 32 filters and a kernel size of 3.

TensorFlow

conv_layer = tf.keras.layers.Conv2D(filters=[1], kernel_size=3, activation='relu')

Drag options to blanks, or click blank then click option'

A32

B64

C16

D128

Attempts:

3 left

💡 Hint

Common Mistakes

Choosing a filter number too small to capture enough features.

Confusing kernel_size with filters.

The number of filters defines how many feature detectors the convolutional layer will learn. Here, 32 filters are specified.

2fill in blank

medium

Complete the code to apply a convolution operation on input tensor x using the conv_layer.

TensorFlow

output = conv_layer([1])

Drag options to blanks, or click blank then click option'

Afeatures

Binput_data

Cimage

Dx

Attempts:

3 left

💡 Hint

Common Mistakes

Passing a variable that is not defined or unrelated to the input.

Using the layer without calling it as a function.

The convolutional layer is applied to the input tensor x to produce the output feature maps.

3fill in blank

hard

Fix the error in the code to correctly define a Conv2D layer with a 5x5 kernel size.

TensorFlow

conv = tf.keras.layers.Conv2D(filters=64, kernel_size=[1], activation='relu')

Drag options to blanks, or click blank then click option'

A5

B(5, 5)

C[5, 5]

D5x5

Attempts:

3 left

💡 Hint

Common Mistakes

Using a string like '5x5' which causes a syntax error.

Using a list instead of a tuple.

The kernel_size parameter expects a tuple for height and width, so (5, 5) is correct.

4fill in blank

hard

Fill both blanks to create a convolutional layer with 16 filters and a stride of 2.

TensorFlow

conv_layer = tf.keras.layers.Conv2D(filters=[1], kernel_size=3, strides=[2], activation='relu')

Drag options to blanks, or click blank then click option'

A16

B1

C2

D32

Attempts:

3 left

💡 Hint

Common Mistakes

Using stride 1 which does not downsample.

Confusing filters with strides.

Filters are set to 16 and strides to 2 to downsample the input by skipping pixels.

5fill in blank

hard

Fill all three blanks to create a Conv2D layer with 64 filters, kernel size 3, and 'same' padding.

TensorFlow

conv = tf.keras.layers.Conv2D(filters=[1], kernel_size=[2], padding=[3], activation='relu')

Drag options to blanks, or click blank then click option'

A32

B3

C'same'

D64

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'valid' padding which reduces output size.

Using wrong kernel size format.

Filters=64, kernel_size=3, and padding='same' keep output size equal to input size.