Processing & Properties of Floor and Wall Tiles.pptx
Cat and dog classification
1. Application to classifying the images in
convolution neural network
Prepared by:
Omar AL-DABASH
Cukurova University
Computer Engineering Department
2. OUTLINE
Deep learning
Convolutional Neural Networks
The problem space
How can the computer recognize images
Our work
3. Deep Learning
Deep Learning is a new area of Machine Learning research, which has been
introduced with the objective of moving Machine Learning closer to one of its
original goals to artificial Intelligence.
The main aim of this learning is to help to achieve and understanding the data such
as images, text and video to recognize them.
4. Convolutional Neural Networks
Most of large companies uses this kind of deep learning at the core of their
service. Facebook uses neural nets for their automatic tagging algorithms,
Google for their photo search, Amazon for their product recommendations,
and Instagram for their search infrastructure.
However, use case of these networks is for image processing.
5. The problem space
When a computer sees an image (takes an image as input), it will see an
array of pixel values. Depending on the resolution and size of the image.
let's say we have a color image in JPG form and its size is 480 x 480. The
representative array will be 480 x 480 x 3. Each of these numbers is given
a value from 0 to 255 which describes the pixel intensity at that point.
The computer is able perform image classification by looking for low
level features such as edges and curves, and then building up to more
abstract concepts through a series of convolutional layers.
6. Our work
Dataset consist of three section
1- Training consist of:
- 4000 of images cat.
- 4000 of images dog
2- Test section consist of:
- 1000 of images cat
- 1000 of images dog
3- 4 images of single prediction
Perhaps we put four images in single predication to testes the system
learned or not.
7. Deep Learning Basics
Deep Learning – is a set of machine learning algorithms based
on multi-layer networks
OUTPUTS
HIDDEN
NODES
INPUTS
Deep learning is also know as a deep structured learning or hierarchical learning. It’s appear in the beginning of 2006 as a new researches area to machine learning. The techniques of deep learning researches have developed over the past years and have influenced a wide range of worth on the information of signal processing in its traditional and modern form. This is within the broad are that include the basic concepts of artificial intelligence and machine learning.
It's a dual branch between biology and math. where this science was one of the latest innovations affecting in the field of vision computer science.
where in 2012 which these networks have grown prominently, where it was used by Alex krizhevsky. He won a prize in ImageNets because he dropped the classification error rate from 25% to 15% where, that was the best improvement at that time.
When you put a photo with your friends on social media sites, it recognize your friends automatically.
The Skype program also translates conversations by individuals at the same time in very high quality.
Behind all these applications is a kind of learning called deep learning.
For humans, the task of recognition is one of the first skills we learn from the moment we are born and is one that comes naturally. Humans have the ability to easily distinguish objects and recognize them without thinking twice.
let’s think about how the computer can approach this. What we want the computer to do is to be able to differentiate between all the images it’s given and figure out the unique features that make a dog a dog or that make a cat a cat.
A more detailed overview of what CNNs do would be that you take the image, pass it through a series of convolutional, nonlinear, pooling (downsampling), and fully connected layers, and get an output
Image classification problem to train CNN if the image is a doge or a cat.