Topic: Convolution Neural Network: Presented by

You might also like

Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 13

Topic: Convolution Neural Network

Presented By:
Fazal Ahmad

M Abdullah Omama Munir

Ayesha Siddiqa Iram Akram


Before understanding the CNN let’s understand the Two important terms
 Image processing
Image processing is a method to perform some operations on an image in order to get an
enhanced image or to get some useful information from it.
 Computer Vision
It is a field of computer science that works on enabling the computers to see, identify and
process image in the same way that human vision does and provide appropriate output.
What is CNN?
 CNN is another type of neural network that can be used to enable machine to visualize things
and perform tasks such as image classification, image recognition and object detection.
 CNN are widely used in the image recognition, image classification, object detection and face
recognition.
Suppose we have given an image of anything

Picture size of 480 x3 CNN Output Label

Like CNN Is classifying the image into 3 classes Dog, Cat, Mouse. Then 1st probability
will be of Dog 2nd will be of Cat and 3rd will be of Mouse.
Example
 Suppose we want to recognize this number to computer , Computer use it RGB numbers
from 0—255.
The issue with this is it’s a very hard coding if we shift a little the number and it’s
representation of the number is changes it will not match with the original number &
computer will not recognize it easily.
How we humans Recognize things
Neurons
Steps of CNN
 Input an Image
In the First we input an image and then apply the further layers of the CNN
 Convolution layer
Here are multiple convolution layers available, every layer abstract a feature . Like in
the image one layer can be used to detect it’s vertical edges, The second one can be
used to detect the objects, and 3rd one can be used to detect Objects shapes
 Max Pooling
Max pooling is an another building block of CNN its function is to progressively
reduce the spatial size of the representation to reduce the amount of the parameters
and computation in the network, Pooling layers operates on each feature map
independently. It reduce the size of pixels , and detect only the important features of
the image.

 Flattening
Flattening is converting the data into 1-Dimension array for inputting It to the next
layer We flatten the output of the convolution layer to create a single long feature
vector & convert the each row into the column. and it is connected to the final
classification model, which is called a fully connected layer.
 Fully Connected
The fully connected layers means that the every neuron in the layer is fully connected
to each other . so that it can process the each data properly . The main aim of this layer
is to use the high level features of input image produced by the convolution and
pooling layers for classification into various class on the base on the training data sets

 Output
The output layer used the soft max activation function , and the soft Max activation
function takes the input in the form of vector and also produce the output in vector.
Where there are the probability value are 0 or 1 to predict about the final decision of
image
Uses Of CNN in Real World Applications

 Face detection:
CNNs have been used to detect faces within images. The network takes an image as
the input and produces a set of values that represent characteristics of faces or facial
features at different parts of the image. CNN has shown improved accuracy over
previous algorithms, identifying faces 97% of the time according to various research
papers.
 Facial emotion recognition:
CNNs have been used to help distinguish between different facial expressions such as
anger, sadness, or happiness. CNNs can also be adapted to perform well with various
lighting conditions and angles of faces within images.
 Object detection:
CNN has been applied to object recognition across images by classifying objects
based on shapes and patterns found within an image.
Uses Of CNN in Real World Applications

 Self-driving or autonomous cars:


 Auto translation
 Next word prediction in sentence:
 Handwritten character recognition
 X-ray image analysis:
 Visual question answering:

You might also like