|
Description: We will implement the algorithm from "Genetic Programming for Articulated Figure Motion". The Genetic Algorithm code will be written for my GA class in Lisp. Chris Klaiber is on my team in GA to help implement this code. The physically based simulation will be the majority of the code to count for Animation Algorithms credit. It will be written in C or C++. We're planning on defining some protocol to talk between the languages with sockets. Lisp is ideal for the GA code. C/C++ is ideal for the simulation code. A more indepth look at the system is available in thr project linked to directly following this paragraph: GA/AA Project Proposal The focus in the genetic algorithms course is experimentation, so we should have lots of good experimental results about optimal settings for performing this process efficiently. Work Plan: Weeks 1 - 3: Planning Break: Beginning Simulation Code Weeks 4 - 6: Final Simulation Code / Beginning GA Code Weeks 7 - 8: Final Code Week 9: Testing Week 10: Final Presentation API: OpenGL Hardware: This is not a real-time problem, so it should run on any hardware, just slowly. Lisp: CMUCL Final Presentation: The final presentation will consist of a PowerPoint presentation that has videos embedded of visual project results. If there's a simple enough test case, I would like to present it in real-time, but this may be ambitious. I'll leave this up in the air until the mid-quarter update. Deliverables: Code Tarball Final Report A Horrible Individual That Was Evolved A Mediocre Individual That Was Evolved A Good Individual That Was Evolved |