Artificial Intelligence with Microsoft Power BI

Chapter 9. Image Tagging

A picture may be worth a thousand words to us humans, but to a computer, a picture provides just a few descriptions, if that. In fairness, image data is incredibly complex to a machine, and it’s taken many advances in artificial intelligence to get where we are today. While computers may be lacking in picture recognition, they excel in their speed and scale of analysis. The prevalence of image data means businesses can unlock new insights by leveraging this AI ability.

By the end of this chapter, you will learn the fundamentals of computer vision and its relationship to artificial neural networks. You will also see examples of how organizations leverage object recognition for problem solving and data-driven analytics. Finally, you will be familiar with the benefits and limitations of image tagging within Power BI and confidently apply it to a real-world dataset of AirBnB listings for building a novel report.

Images as Data

We live in a time when data is cheap and getting cheaper each year. In 1956, 1 MB of disk storage cost over $100,000 in today’s dollars. Now the same amount of disk storage costs 1/667th of a cent. This massive depreciation has allowed businesses to affordably store large amounts of data either on premises or in the cloud. This includes many different types of data, from numbers and text to more space-intensive formats like audio and video. These multimedia file types provide mind-boggling amounts of data for business use. Statistica ...