Hello everyone, welcome to a course on KNN image classification. So, in this course, we will learn how to use machine learning to classify images using Kanan and again N stands for K nearest neighbor is one of the algorithm that is used for image classification. And this course is made for absolute beginners who are new to the machine learning world and wants to get a handle on how to train a machine to classify images. So, in this course, we are going to learn to classify images. Basically we are going to train a machine to learn on how a machine can classify images to the advantage of that is you know, human beings can classify images like if you are looking at video of a dog, you can say a dog then you if you see a video of cat, you can say we have cat But there's a limit to human potential like you cannot do that at scale, right, you have to have humans doing that task, they have to watch those images for a long period of time.

And if you want to classify millions and billions of images, I think that's too expensive and takes a lot of time for humans to do that. So it would be better if a machine can do that right machine once the machine when you can train a machine on how to do that you can really classify millions and billions of images in a matter of minutes. So you can achieve a huge scale with the help of this algorithm. Also, once the computer is trained, you can use it in various applications like if you want to have billions and millions of images. And if you want to classify them and put into appropriate folders, you can use that you can also help classifying image in To labels like if you are image of a cat, then you can label it as a cat. Similarly, image of a dog, you can label as a dog.

And you can store that information along with the images so that when, when the search engine search for your images, you can also use those textual tags and you can find your images page to the actual tax write. And you can also translate those images, the text into other languages, so that people can from other languages can find the images. For example, if there's an image of a dog, which is created by an English user, like me, and if and it can classify this image and put a label as a dog, and I can translate translated into French, like in French, you call dog a shin. And if, and you can store that also along with the image when the when the French user comes in, and if he's doing a search for Shin, you can still find the image right. So text to image Translation becomes very easy when you classify images.

We will also learn how machines perceive image. They don't perceive it the same way as humans do. And that presents its own set of challenges and we will see several techniques to address those challenges. We will also learn how you can encode image as a vector, because then you can use this vector to find similarity. And we will also learn about different algorithms, which are used for identifying the image similarity. Then, I will also walk through a code where we will train our train a machine using Kanan algorithm.

And then we will classify the images into the actual labels and we will also measure how accurate our algorithm is. This course is made for absolute beginners who are completely new Machine Learning world and if they have just you just need to have a very basic knowledge of Python. And I will explain you each and every technique that is used for classifying images. So, if you are new to the machine learning world and if you just want to understand how you can learn computer vision, this is a great introductory course. Now, the algorithms that are used in this introduction like that re course are meant for beginners. Basically, we are using a K nearest neighbor algorithm that is very highly simplistic algorithm that is used for unders from and the goal here is to emphasize the understanding of the algorithm and not creating an algorithm which has high accuracy rate 90 and 85%.

Gain in although is a very basic algorithm. It doesn't give you a very high accuracy. But it's a great way great foray into machine learning just to understand how machine perceives imagine how They can classify images. The more popular Gotham's are like computer convolution neural networks, but they are more difficult to understand. So this core The goal of this core is make sure you have a basic understanding of image classification, how much you miss ncma and some basic algorithms. This course is not meant for advanced users who really want to understand and take the competition to the next level that will be covered in some other course.

So just want to make sure I set realistic expectations. So this course is a great for beginners and I will I am trying to make sure I remove all the I explained the all difficult part by making it as easy as possible. So so let's go for it.