This sketch uses an evolutionary algorithm to refine the neural network which drives the critter. The fitness of a creature is the number of food particles eaten after 1000 ticks.
As inputs, the critter receives the distance from each of its antennae (dark red circles) to the closest food particle. It outputs 2 values, one indicating the desired turning force, and one indicating the desired forward force.
Left Click -- Zoom In
Right Click -- Zoom Out
Mouse Move -- Change Perspective
This is the final installment in my Bubbles series. Because of the dependence on the controlP5 library, I had to upload it as an applet. For my most recent HTML5 Canvas version, check out my Bubbles 2 sketch here: http://www.openprocessing.org/visuals/?visualID=49296
This sketch came out of some experimentation I did with parallax scrolling. I came up with my own radial blur system, though I have since found other projects using similar techniques. If you're interested in finding out more about the evolution of this project, an earlier version can be found here: http://www.openprocessing.org/visuals/?visualID=49284