Teaching Robots Through Infant Imitation

04 Dec Teaching Robots Through Infant Imitation

Researchers from Washington University are teaching robots to learn much like infants do by experience and imitation rather than by traditional programming. In one experiment a physical robot learns to focus it’s gaze where a human is looking by tracking the beginning and end point of the humans gaze. “Babies learn through their own play and by watching others,” says Meltzoff, “and they are the best learners on the planet — why not design robots that learn as effortlessly as a child?”

Journal Reference:

1. Michael Jae-Yoon Chung, Abram L. Friesen, Dieter Fox, Andrew N. Meltzoff, Rajesh P. N. Rao. A Bayesian Developmental Approach to Robotic Goal-Based Imitation Learning. PLOS ONE, 2015; 10 (11): e0141965 DOI: 10.1371/journal.pone.0141965

Read the Full Article Here