Pokemon Go and IBM Watson: When AI Meets AR

Pokemon Go almost immediately after the release became an extremely popular game. Thanks to this application, millions of fans of mobile games began to walk just huge distances per day. Some Pokemon Go enthusiastic users perform and exceed their normal standards in the number of steps taken per day.

Finding some Pokemon can be very difficult - and what kind of game would it be if it could be completed in a few hours? Pocket monsters of various kinds have to search for weeks, and not always such searches are successful. Therefore, a number of developers began searching for ways to make life easier for players in Pokemon Go. A San Francisco developer named Michael Hsu got the IBM Watson cognitive system to look for.

Michael took part in the AT & T Shape Tech Expo hackathon and created a useful application for many Pokemon Go fans. He himself was very interested in pokemon. By his own admission, he is a “pocket monster hunter” since the release of the application.
')
To simplify the task of searching for animals on the map, Su began to look for an API that is available and suitable for solving his task. Almost immediately, the developer saw the new Watson Visual Recognition API, and decided to use it for his own purposes.



Su created an application that periodically takes screenshots in the Pokemon Go application. Watson Visual Recognition API is used for image analysis. If the pokemon is in the screenshots, the cognitive system reports this. The system collects this information and notifies players in Pokemon Go about where Pokemon certain species are located.

According to Stefanie Kachmarchik, a representative of the IBM Digital Group, the new application shows how easy it is to use the possibilities offered by the cognitive system. “You have had the opportunity to use the results of many years of research and development of our company as a service, when the user can say:“ I have a photo of what it is? ”, Says Kachmarchik. The application using the Watson Visual Recognition API was created in less than 24 hours. Without this service, developing such an application would take much longer.

“Watson can track Pokemon around the world, and players will see where and what Pokemon are hiding,” Kachmarchik comments on the program’s capabilities.

Soon, the source code of the application will be uploaded to GitHub , and other developers will be able to familiarize themselves with it. If desired, you can modify the application.