The most common use of Artificial Intelligence is object recognition. Humans take object recognition for granted, but robots need some help. The main reason that most robot competitions use human drivers is the robot’s lack of knowledge of its surroundings. VEX AI uses software and neural networks that can detect the game pieces, goals, and objects that VEX robots want to interact with.
How VEX AI Vision System Works
Many object classification systems are designed to determine the one subject of the image with some certainty, such as a cat with 96% certainty. The VEX AI system is designed to identify hundreds of instances of various objects within an image and locate their screen coordinate. In addition the VEX AI system measures distance to each object using stereo vision. That information is combined, along with the position information from the VEX GPS to locate each object in 3D space.
How it Applies
Writing code to sort through the data and make informed and strategic decisions is a key part of Computer Science and competitive autonomous robotics.