Intro To Convolutional Neural
Networks
Presenters :
Basit Rafiq
Sunil Ashraf
 Convolutional Neural Network
 (CNN) is a type of artificial neural network or
algorithm used in image recognition and processing
that is specifically designed to process IMAGES.
Why CNNs?
CNN Model
Convolution
This Process is the
multiplication of the
given image matrix
with the required
filter.
It makes the input of
our next step pooling.
Convolution Process
Pooling process
Pooling. Its function is to progressively
reduce the image size of the
representation to reduce the amount of
computation in the network.
Types
1) Max Pooling
2) Average Pooling
3) Sum Pooling
Pooling Calculation
Max Pooling
Results
Traditional Approach To
Image
Classification
Input Image preprocessing Classifier Object Label
https://papers.nips.cc/paper/4824-‐imagenet-‐classification-‐with-‐deep-‐convolutional-‐neural-‐networks
Image Classification
Generate An Image From A Sketch
https://affinelayer.com/pixsrv/
Part of Preprocessing (rotation)
Object Recognition
• This animal is a combination of
arms, legs, & a tail in specific
proportions.
• Similarly recognizing trees,
grass..etc (place)
• Object recognition process
Image Features
• Image recognition is something recognizing
the different angles of the image.
• Edges
• Lines at different angles, curves, etc.
• Colors, or patterns of colors etc..
CNN Layer Architecture
Pooling (optional)
Labeling
Convolution
Input (Image)
Ideally We’d Learn Features
Input
Image
CNNs
Output
Label
Applications of CNN
1.Decoding Facial Recognition
2.Analyzing Documents
3. Environmental Collections
4.Advertising
5.Recognizing Images
6.Many more
Questions
?

Convolution Neural Network (CNN)