Nature of Code Final Proposal: Food Chain Game

For my final project, I’m taking an idea from last week’s assignment, and expanding on it to incorporate everything we covered in the class. 

Last week, I used the ml5 image regression example to make a ‘game’ that forces the user to make some really unstable-looking faces to be categorized as “emotionally stable”


I like the idea of making the user interact with the animated p5 canvas, so I want to make a game using the poseNet to get the user to interact with particle systems on screen.

I’m taking some inspiration from two examples we saw in class:

Steering Agent by @filiplazovic

Eat, Mate, Die by Sam Levigne

As well as from a game I used to play when procrastinating called “Fishy” 

In fishy, you play an orange fish who needs to eat all the fish that are smaller than it while avoiding any fish who are bigger than you.  There are no levels, but the bigger you get, the more fish you can eat, and the harder it is to avoid any predator fish.


For this game, I want the user to look at their own reflection in the camera, and interact with some particles that will animate over the video capture image. One set of particles will function as food that replenishes your health meter, and one will act as predators that harm you and diminish your health meter. The longer you stay alive, the higher your score!

I havent decided yet whether to make both escaping the predators and catching the prey harder as time goes on, or whether, like in fishy, I’ll make it so that food is easier to catch as you go on. However, I think I want to use a genetic algorithm with at least one of the populations to make it increasingly harder as time passes to catch food and stay alive!

Leave a Reply

Your email address will not be published. Required fields are marked *