Training AI to play Space Invaders against an authentic version, using an alternative AI approach.
Rudimentary vision recognition used to steer an explosive drone boat into a target.
Training bees to collect nectar and returning it to the hive.
Pong 4.0, an attempt to use a "vision" approach to track the ball.
Can ML remember multiple images within a single network?
AI experiments on training virtual life forms.
Using ML to drive blobs, go-karts and tanks.
Is it possible to train AI to recognise font drawn digits with 100% accuracy?
Rather than fix potholes, can we train the cars to dodge them?
Creating a more advanced Pong AI that learns as it plays.