Applications of AI

- [Instructor] One of the most common applications of AI, image classification and object detection in images. Maybe the most common application that's linked to everyday life. If you have a smart phone that has facial recognition, it contains elements of this technology. If you upload images to major social media platforms, objects maybe inferred in photos in order to help the tagging and sorting of the images. Object detection has grown in popularity over the last decade, thanks to the increase in computing power, access to larger curated datasets and advances in the architecture of neural networks. One of the largest created datasets initially was the Imagenet dataset. Which contains thousands of tagged images that can be used to create classification models. The actual story behind image that is quite amazing and the initial creators pushed really hard to bring that source to life. Since then, it's been updated yearly and used for countless competitions. The actual technical training of neural networks is out of scope for this course, but it's important to know that these three elements have waved the strides in innovation. As mentioned, the actual technical training aspects are out of scope for this course. But these tools have been available on your own iPhone and on different MAC computers through X code, which is a free software solution. It has a very simple drag and drop interface. And all you need are the images that tags for those images and a specific outcome that you'd like to build towards. Since the actual technical training of networks is out of scope for this course, I should mention that to actually train networks is easier than it's ever been. X code on Apple MAC computers provides a drag and drop interface for you to be able to design your own models and deploy them on an iPhone. All you really need is a number of images and tagging and to follow a few simple steps to train and then deploy a model. The remaining other examples of using similar structures and techniques to predict different outcomes across various industries. There's a constant flow of research and new applications are being shared across the web in both private and public institutions. Although image classification and object detection may be some of the most visual applications, the same tools that is being used across many different applications. Such as audio recognition and sound generation. Identifying instances of fraud in large numbers of financial transactions. Estimating structure and movement of a body over time and predicting what task will be done. These types of applications are only really scratching the surface. And as we progress through the course, we will see how these applications are used in different ways. Sometimes with, or without the consent of the original data providers.