10 Questions
What is the size of the input before being passed to a fully connected layer?
8×8×64
If the input is flattened before being passed to a fully connected layer, what will be the size of the flattened output?
4096
What type of layer typically follows a flattened input?
Fully connected layer
What is the purpose of flattening the input before passing it to a fully connected layer?
To convert the input to a one-dimensional structure
What would be the size of the flattened output if the input size was 4×4×64?
1024
What is the significance of the number of filters in a CNN filter?
It determines the number of activation maps produced by the filter
What is the purpose of the channels (depth) dimension in a CNN filter?
To specify the number of input features in the data
What is the size of the receptive field in a CNN filter with a size of 5×5×3×10?
5×5
What does the height and width of a CNN filter specify?
The size of the receptive field
How many activation maps will each filter produce in a CNN filter with a size of 5×5×3×10?
10
Study Notes
Flattening Input for Fully Connected Layer
- Given an input with size 8×8×64, when flattened, the output size will be 4096.
- This is because the input dimensions are multiplied together to get the total number of elements.
CNN Filter Characteristics
- A CNN filter with a size of 5×5×3×10 produces 10 activation maps.
- The filter size is broken down into height (5), width (5), channels (3), and number of filters (10).
- Each filter generates a separate activation map, resulting in 10 maps.
This quiz question assesses understanding of how neural network layers process data. It asks about the size of a flattened output from a 3D input before passing to a fully connected layer.
Make Your Own Quizzes and Flashcards
Convert your notes into interactive study material.
Get started for free
