CNN Classic CNN Architectures 5

Question 1

LeNet-5 uses which type of pooling in its subsampling layers?

Accepted Answer

Average pooling

Answer

LeNet-5 uses average pooling (called 'subsampling') in its pooling layers, which was the standard before max pooling became dominant.

Question 2

What key observation about 3x3 convolutions did VGGNet exploit to justify using only small filters?

Accepted Answer

Two stacked 3x3 convolutions have the same receptive field as one 5x5

Answer

Two stacked 3x3 conv layers cover the same 5x5 receptive field but with fewer parameters (18 vs 25 per channel) and an extra nonlinearity.

Question 3

GoogLeNet includes auxiliary classifiers during training. What is their primary purpose?

Accepted Answer

To combat vanishing gradients by injecting gradient signal at intermediate layers

Answer

Auxiliary classifiers in GoogLeNet inject gradient signal into earlier layers during backpropagation, helping train the deep network by fighting vanishing gradients.

Question 4

Which classic CNN was specifically designed to fit on embedded systems with a model size under 0.5MB using 'fire modules'?

Accepted Answer

SqueezeNet

Answer

SqueezeNet (2016) used 'fire modules' with squeeze and expand layers to achieve AlexNet-level accuracy at less than 0.5MB, targeting embedded deployment.

Question 5

What is a 'projection shortcut' in ResNet, as opposed to an 'identity shortcut'?

Accepted Answer

A shortcut that uses a 1x1 convolution to match dimensions when channel sizes change

Answer

A projection shortcut uses a 1x1 convolution (with stride) to match the spatial and channel dimensions when the skip connection would otherwise have a size mismatch.

CNN - Convolutional Neural Networks Practice Test

CNN - Convolutional Neural Networks Practice Test

CNN Classic CNN Architectures 5