Difference between Computer Graphics and Computer Vision | Relationship with AI, ML, Deep Learning
Hello everyone and welcome to another video blog on the computer graphics video lecture series
and today we are going to cover how different fields are interlinked with each other which are like computer graphics, computer vision, artificial intelligence, machine learning, and deep learning.
Let's start with the computer graphics, it is an important field where the computers use to take an input of any defined model algorithm textual information or even graphical information or data and it used to do some processing and produce an improved or enhanced is a new graphical visualization.
Next is computer vision, where the computer is used to take input of an image or graphical data and it's used to produce some sort of information that can be used for the decision making analysis or to do some sort of further processing. In computer graphics, the output always used to be another graphical thing or a new graphical model or design, while in the computer vision the input used to be graphical data in the form of video or visual information images, and the output used to be some decision information. So, both fields are almost opposite to each other in terms of producing the output and in terms of taking the input.
While, how these two domains are linked with the other domains like artificial intelligence, machine learning, and deep learning before going to that part I will discuss these three fields.
Artificial intelligence is the main field where tries to make the computing device much powerful that they will start acting like the human brain and they will start making the decisions on their own based on the trained models or based on the set of rules.
Machine learning is a subfield of artificial intelligence where we use to let the machines learn the way our brain used to learn like at the start, we used to train our brain by feeding some sort of information. In the same way, we let the computers get trained by giving a large amount of data, and then we used to give some new data in order to produce the output for analysis or decision making.
Deep learning is a subfield of machine learning where the prediction rates are much higher as compared to the other typical machine learning algorithms. By using deep learning algorithm, the input training data need to be extremely high and we used to let deep learning algorithms extract the features by their own and then use those features to take the decision.
The artificial intelligence has sub-field machine learning and machine learning has sub-field deep learning. Now how these three domains are interlinked with the computer graphics and computer region, if we see in computer graphics, we use to let the computers generate new graphics and how we can generate those graphics? Of course, we need to have some sort of algorithms running in the background. One option is we can generate graphics by direct user interaction with graphical software and doing the arrangements of different graphics and doing some processing by using photoshop or AutoCAD. The other option is that we can use the programming algorithms and we will define a set of rules and the new produced model will be acting based on those defined set of rules or defined models or defined algorithms. Here the idea of artificial intelligence and machine, learning and deep learning comes, where we can generate the new graphical images by applying these techniques.
I can give you several examples like we can do image enhancement and we can remove the noise from the images by using artificial intelligence and machine learning algorithms. In the same way, in order to extract different objects from the images or from the graphics, we can again apply artificial intelligence and machine learning. Another example like we want to merge our combined different objects or different graphics together to act collectively in a single new graphical image or graphical representation. In this case, again we can apply artificial intelligence or machine learning.
If we will apply machine learning and artificial intelligence, manual work will be reduced and even we can get better representation if we will have a better-defined model for the new graphical images. Therefore, we can say, all these fields are interlinked somehow because all of them are using computing devices in order to generate some sort of representation or results or in order to take decisions similarly.
For the computer version, a graphical image is used to be fed to the computer in order to produce some sort of decision making, recognition, or prediction. So, in that case, again artificial intelligence is the one which is actually used at the backend and the prediction are being done by using the rule-based methods or by using the machine learning algorithms.