Teem

The next generation open evolutionary framework

See also: Download, Enki, Ishtar, An

Stephane Magnenat, Antoine Beyeler, Markus Waibel and other contributors

Introduction

Teem is a software framework that allows for easy creation and execution of evolutionary robotics experiments.

It can be combined with a variety of libraries and programs to form a powerful platform for anything, from evolution of simple Braitenberg behaviour to complex steady state evolutionary experiments involving groups of robots.

Teem is free software, published under the GNU General Public License with the restriction of credit to authors mentioned below and in the documentation.

Modules

Teem provides several modules that can be used alone or linked together.

	Genomes Linear, bit based Tree-like, dynamic topology, float based Evolutions Classical GA Steady-state GA
	Neural networks topologies Feedforward, including backpropagation and recursion Fully connected Neurons types Sigmoidal Spiking, including STDP
	Sample experiments Evolution of obstacle avoidance Evolution of vision-based navigation Evolution of homing

Teem and its dependencies

Teem depends on the following libraries, which are also available on this site:

• An: A 2D scientific toolbox (mandatory)
• Ishtar: A network toolbox (mandatory)
• Enki: A fast 2D robot simulator (optional)

Figure 1 shows the interaction diagram between Teem and those libraries.

Figure 1: Teem interaction diagram with the libraries it depends on.

License

Teem is free software released under the GNU General Public License with the only restriction that any publications presenting results obtained with Teem must mention it and its authors.

Download

• Version 1.0: teem-1.0.tar.gz (documentation)
  
  • depends on: liban-1.0.tar.gz (documentation)
  • depends on: libishtarnet-1.0.tar.gz (documentation)
  • depends on: libishtarservices-1.0.tar.gz (documentation)
  • depends on: libishtarsettings-1.0.tar.gz (documentation)
  • depends on: libboost (external)
  • optionally use: libenki-1.0.tar.gz (documentation)
  • optionally use: libSDL1.2 and libSDL-image1.2 (external)
  • optionally use: libavcodec (external)

Last update: Stephane Magnenat, 8/8/2005