Inspiration
We wanted to learn how to implement machine learning to create interesting obstacles for the player and challenge ourselves in learning this novel technology.
What it does
It's the popular Tron-Lightcycles game with an AI opponent that utilizes machine learning to adapt to the player's choices.
How we built it
We used Godot, a popular game engine, to build the game itself. After-which we looked into using a Godot C# based library to provide machine learning functionally.
Challenges we ran into
Machine learning and artificial intelligence can be extremely complex.Quite often the documentation can be sparse and implementation only successful after much trial and error. Not to mention that due to the nature of machine learning it's not often certain whether an issue with AI functionality is due to faulty code implementation or a lack of training.
Accomplishments that we're proud of
We're proud of being able to create an opponent that 'thinks' for itself and of taking our first steps with this emerging technology.
What we learned
We learned a little more about the terminology in the machine learning space. Improving our skills with both Godot and Pytorch.
What's next for AI Light Cycles
There are a variety of features we would like to add, such as a better AI Agent, better graphics and more AI opponents.
Built With
- godot
- python
Log in or sign up for Devpost to join the conversation.