CNN - Convolutional Neural Networks Classic CNN Architectures Questions and Answers

Question 1

The VGGNet architecture is renowned for its simplicity and depth. What is the defining characteristic of its convolutional layers?

Accepted Answer

It exclusively uses very small 3x3 filters stacked on top of each other.

Answer

It uses large 11x11 and 5x5 filters in the initial layers.

Answer

It introduces parallel 1x1, 3x3, and 5x5 filters in each block.

Answer

It employs skip connections to bypass certain convolutional layers.

Question 2

A key innovation of the GoogLeNet (Inception v1) architecture was the 'Inception module'. What is the primary purpose of this module?

Accepted Answer

To perform convolutions at multiple scales (1x1, 3x3, 5x5) in parallel and concatenate the results.

Answer

To drastically increase the network depth by using residual connections.

Answer

To simplify the architecture by using only 3x3 convolutions.

Answer

To replace all fully connected layers with global average pooling to reduce parameters.

Question 3

Which of the following classic CNN architectures was the first to successfully demonstrate that significantly deeper networks could be trained by using 'skip connections' or 'residual blocks' to address the vanishing gradient problem?

Accepted Answer

ResNet

Answer

AlexNet

Answer

VGG-16

Answer

LeNet-5

Question 4

An engineer is choosing a classic CNN architecture for a task on hardware with limited computational resources. They need a model that is deep and accurate but has a significantly lower parameter count compared to VGG-16 or AlexNet. Which architecture would be the most suitable choice?

Accepted Answer

GoogLeNet (Inception v1)

Answer

LeNet-5

Answer

AlexNet

Answer

VGG-19

Question 5

Which of the following was a key innovation introduced in the AlexNet architecture that distinguished it from its predecessor, LeNet-5?

Accepted Answer

The use of the Rectified Linear Unit (ReLU) activation function.

Answer

The use of average pooling for subsampling.

Answer

The introduction of the Inception module.

Answer

A shallow architecture with only 5 layers.

Question 6

The LeNet-5 architecture, a pioneering CNN, was primarily designed for a specific task and had characteristic features for its time. Which of the following best describes LeNet-5?

Accepted Answer

A shallow network using average pooling and tanh/sigmoid activations, designed for handwritten digit recognition.

Answer

A deep 19-layer network using only 3x3 filters for large-scale color image classification.

Answer

A model that introduced residual connections to train networks over 100 layers deep.

Answer

An architecture that used parallel multi-scale convolutions and global average pooling.