Let's see how we teach Artificial Intelligence
To start, a simple learning to visualise how the robot is able to take decisions by itself and to improve them by learning from experience.
The objective here is to maximize the mean robot's speed.
Adding a simple sensor will drastically improve the robot's strategy, the ultra sound sensor give a distance value to its environnement.
The robot, with this new information, is able to learn to avoid obstacles.
The camera provides a much more dense information for the robot, it will take longer for him to learn to avoid walls, but with much more efficiency.
With a little imagination, we can create a lot of new objectives for the robot to accomplish. In this case, the objective for alphai is to get the most green color in the image of his camera.
The result of this will be alphai shooting and pursuing the green balloon.
19 / 10 / 2019
24 / 11 / 2018
30 / 03 / 2019
05 / 06 / 2019
06 / 06 / 2019
20 / 11 / 2019