Developer uses Machine Learning to train an Android game bot to play Rock Paper Scissors
LG is trying to bring back air gestures with Air Motion on the LG G8 ThinQ, but developer ramtin8731 has a better idea of how to use gestures: mobile gaming. For the really, really bored mobile gamers out there, the developer made a Rock Paper Scissors Android game with an added twist: it uses your phone’s front-facing camera to capture your hand movements so you don’t ever have to touch the screen. Better yet, the computer opponent learns from your patterns as you play.
The app will watch what gestures your hand makes, whether it be a rock, paper, or scissors, and let you play against a bot. The app can work offline and takes advantage of TensorFlow Lite to learn your hand gestures and playing pattern. The app’s description says it needs a front-facing camera, obviously, and also a fairly powerful processor to handle the AI processing. In the settings, you can enable the use of the Android Neural Network API which takes advantage of the AI processing capabilities of many modern chipsets.
You can download the app from the Google Play Store link at the end of the article, or visit the forum thread linked below to give feedback to the developer.
Here’s a video shared by the developer showing the app in action: