1. A Presentation
on
Convolution Neural Networks
SESSION 2019-20
Apex Institute of Engineering and Technology, Jaipur
Department of Computer Science & Engineering
Submitted to: Presented by:
Mr. Mohit Saxena rakeshsaran
HoD CSE Department 16EAXCS
4th Year (CS)
13 April 2020 Presented by :- Amar Tiwari
2. Contents:
• How a Computer Reads an Image ?
• Why not Fully Connected Networks for Image Recognition?
• What is Convolution Neural Network?
• How Convolutional Network Works?
• Convolution.
• ReLU.
• Pooling.
• Fully Connected.
• Convolutional Neural Network Use-Case .
23 April 2020 Presented by :- Amar Tiwari
3. How a Computer Reads an Image ?
33 April 2020 Presented by :- Amar Tiwari
4. Why Not Fully Connected Networks?
43 April 2020 Presented by :- Amar Tiwari
5. Why Convolutional Neural Network?
• In case of CNN the neuron in a
layer will only be connected to
small region of the layer before it,
instead of all of the neurons in a
fully-connected manner.
53 April 2020 Presented by :- Amar Tiwari
6. What is CNN ?
• CNN is a type of feed-forward
artificial neural network in which the
connectivity pattern between its
neurons is inspired by the organization
of the animal visual cortex.
63 April 2020 Presented by :- Amar Tiwari
7. How CNN Works?
CNN have following layers :
Convolution.
ReLU Layer.
Pooling.
Fully Connected.
73 April 2020 Presented by :- Amar Tiwari
9. ReLU Layer:
In this layer we remove every negative values from the filtered images and
replaces it with zero’s .
This is done to avoid the values from summing up to zero .
Rectified Linear Unit (ReLU) transforms function only activates a node if the input
is above a certain quantity, while the input is below zero, but when the input rises
above a certain threshold, it has a linear relationship with the dependent.
93 April 2020 Presented by :- Amar Tiwari
11. Pooling Layer:
In this layer we shrink the image stack
into a smaller size.
Steps:
Pick a window size (usually 2 -3).
Pick a stride (usually 2)
Walk your window across your
Filtered images.
From each window, take the maximum
value.
113 April 2020 Presented by :- Amar Tiwari
12. Stacking up the Layers :
123 April 2020 Presented by :- Amar Tiwari
13. Fully Connected Layer:
This is the final layer where the actual
classification happens.
Here we take our filtered and shrinked
images and put them into a single list.
133 April 2020 Presented by :- Amar Tiwari