"Jaws" was an exercise in the artificial intelligence course of my masters study program. It is a small game, where the player controls a shark and tries to eat as much fish as possible. The fish movement works like a Braitenberg Vehicle (http://en.wikipedia.org/wiki/Braitenberg_vehicle), where the shark is the light source and the fish tries to escape from it. The shark can be controlled via the arrow keys. The exercise consisted of implementing a kd-tree (http://en.wikipedia.org/wiki/Kd_tree) nearest neighbour search. Other than the native method of finding the nearest neighbor, this algorithm allows for a much larger amount of fish in the tank without the FPS to decrease. Unfortunately it cannot perfectly be seen in this video, due to the usage of free software that made a huge impact on the FPS. Therefore a screenshot is added in the end, that displays the actual FPS. Our goal was to simulate a basic swarm behaviour/intelligence. The fishes try not to touch each other, so they slightly change their direction when intersecting with others. Additionally they try to avoid the shark. If the come to close to it, they swim faster and change their direction until they reach a certain distance, away from the shark. Download the Jaws.jar at: https://dl.dropboxusercontent.com/u/33732599/Referenzen/Jaws.jar Team: Simon Mary, Benjamin Lebrecht, Simon Hohl
