Google’s street view seems to tell us a lot more than you would think. How much do your street view images say about you?
Researchers at Stanford used AI to analyze around 50 million images in over 200 cities across the United States. From this, they were able to uncover a lot of different things. Their results were even able to determine that there is a powerful link between what vehicles people choose and their political views. Areas in which sedan drivers outnumbered pickup truck drivers the city saw an 88 percent increased change in those voting Democratic whereas areas that had the opposite (pickup trucks over sedans) there was an 82 percent chance the area would vote Republican.
Data in this study was also able to uncover things like the racial makeup, levels of income, and even education levels in the areas studied. All of these things were used in their algorithms. Using their knowledge to determine political leanings of a given neighborhood is quite easy and only requires looking at what kinds of cars are on the streets. This technique according to researchers could even someday replace the census surveys.
Fei-Fei Li, an associate professor of computer science at Stanford and director of the Stanford Artificial Intelligence Lab has said as follows:
“Using easily obtainable visual data, we can learn so much about our communities, on par with some information that takes billions of dollars to obtain via census surveys.”
“More importantly, this research opens up more possibilities of virtually continuous study of our society using sometimes cheapy available visual data.”
“It can help us understand how our society works, the things people need and how we can improve lives.”
In order to create the system that made all of this possible the team had to work to build by hand an image database of all cars since 1990 – year, make, model, trim packages and then teach a computer to recognize the subtle differences between them all. Even in images that are at odd angles or in which the vehicle is hard to see. Once everything was said and done their machine was able to sort all of the cars in all 50 million images into under 3,000 categories over the period of fewer than 2 weeks.
Technology truly is capable of so much more than we give it credit for. I for one cannot wait to see what else is uncovered using google street view images. Who knows what interesting things will come to light.