Jump to: navigation, search

Ad Hoc Teamwork: HFO

Revision as of 23:41, 23 October 2014 by Sbarrett (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

This page presents some videos on ad hoc teamwork in the Half Field Offense (HFO) task in the 2D simulated RoboCup domain. The original specification of the HFO task can be found here. An overview on ad hoc teamwork can be found here, and LARG's publications on ad hoc teamwork can be viewed here.

In this research, we investigate the performance of PLASTIC-Policy in the HFO domain. PLASTIC-Policy learns policies for cooperating with past teammates and selects from among these policies on the fly. In these videos, PLASTIC-Policy has learned about 7 previous types teammates. As a baseline, we compare the performance of the Combined Policy, where the ad hoc agent combines its observations about all 7 types of teammates to learn a new policy.

The green offensive agent is the ad hoc agent. The yellow offensive agents are externally-created teammates, meaning that we did not design them. Instead, they are taken from the 2013 RoboCup competition in the 2D simulated soccer league. The red agents are defensive field players while the purple agent is the goalie.

Limited Half Field Offense

In the limited half field offense task, 2 offensive agents play against 2 defensive agents, including the goalie. In these videos, the defense is using behaviors designed by Helios. We sample teammates coming from 2 different teams for the videos.

Cooperating with a teammate designed by Cyrus:

Combined Policy:          PLASTIC-Policy:
Off: 2 Def: 8 Off: 6 Def: 4
Download Download

Cooperating with a teammate from the agent2d code release:

Combined Policy:          PLASTIC-Policy:
Off: 2 Def: 8 Off: 5 Def: 5
Download Download

Full Half Field Offense

In the full half field offense task, 4 offensive agents play against 5 defensive agents, including the goalie. In these videos, the defense is using behaviors from the agent2d code release. We sample teammates coming from 2 different teams for the videos.

Cooperating with teammates designed by Helios:

Combined Policy:          PLASTIC-Policy:
Off: 2 Def: 8 Off: 5 Def: 5
Video Video

Cooperating with teammates designed by the Gliders:

Combined Policy:          PLASTIC-Policy:
Off: 3 Def: 7 Off: 6 Def: 4
Video Video