Researchers at the Carlos III University in Madrid (UC3M) have programmed clones that imitate the actions of humans playing soccer on a computer using behavioral cloning techniques. The clones learned the players' behaviors and applied that knowledge to avoid opponents and score goals. UC3M study author Ricardo Aler says the researchers used behavioral cloning to program a virtual player by observing the actions of a person playing in the simulated RoboCup league. The RoboCup is an international robot soccer championship intended to promote artificial and robot intelligence development. The event's promoters want to develop a team of totally autonomous robots that are capable of beating the best human players by 2050. Aler says there are several leagues within RoboCup, including a league of real robots and a league of simulated robots, which Aler and his team compete in using a software model called Robosoccer. "The human player plays Robosoccer as if it were a video game, and the system observes both the stimuli that the person is receiving from the screen as well as the actions he or she is carrying out on the keyboard in order to shoot or pass the ball," he says. The researchers then use behavioral learning techniques to create a model of how the person played, which is then used to create a "clone agent" that imitates the real player. Using behavioral cloning, the researchers found that the clone agent can score goals against an opponent, much like human players.
View Full Article
No entries found