Source Code and Binaries

3D Simulation League

UT Austin Villa RoboCup 3D Simulation Base Code Release
The 3D simulation team's base code release is hosted on GitHub.
The 3D simulation team's released binaries can be found on the 3D simulation team homepage.

Standard Platform League

UT Austin Villa SPL 2012 Code Release
Source code release for the 2012 season: austinVilla2012.tar.gz
This code release consists of the core of the UT Austin Villa 2012 code. The code should work out of the box, if the color tables match the environment encountered. There are some differences from the code used in the 2012 competition, most notably that the behavior code provided is much simpler.
Though we are providing the code "as is" with no promised support, we are happy to receive feedback, and we encourage people to let us know if they are able to make good use of the code. For any questions, contact UT Austin Villa at UTAustinVilla-legged (at) cs (dot) utexas (dot) edu.
If you use this code for research purposes, please cite this paper.

Legged League

Aibo Kick Learning Source
Source release associated with kick learning research. Includes Tekkotsu classes and machine learning algorithms needed to recreate the experiments in this project.

Simulation Leagues

Austin Villa Coach Binaries
The Austin Villa simulated coach is an agent capable of giving RocoCup soccer agents strategic advice about general play from handcoded rules and opponent-specific advice from rules learned through observing past opponent play.
The 2003 coach was implemented mostly from scratch. However, some of the code used to interface with the soccer server was taken from the UvA Trilearn base client. This is a Linux binary and has only been tested on Debian 3.0.
The 2004 coach was again implemented mainly from scratch. However, some ideas were taken from the 2003 version. This is a Linux binary and has only been tested on Debian 3.1.
Austin Villa Coachable Player Source
The Austin Villa coachable players are RoboCup soccer agents capable of incorporating advice given in the standardized coach language (CLang). The players are built upon the UvA Trilearn base client. This code has only been tested on Debian 3.1.
Valid CSS!
Valid XHTML 1.0!