- Taxes on some wealthy French top 100 pct of income: paper
- North Korea fires short-range missiles for two days in a row |
- Israel warns against Russian arms supply to Syria
- Winning ticket for $590.5 million Powerball lottery sold in Florida |
- Female hostage died from police bullet in New York standoff: official
Emotive robots take lessons from toddlers
PASADENA, Calif. |
PASADENA, Calif. (Reuters Life!) - Keepon, a bright yellow, silicon-skinned, snowman-like robot can sway from side to side with joy, and bob up and down with excitement, much like toddlers.
It may not experience life in quite the same way as humans do, but it can express some of their emotions, however basic, breaking the stereotype of an unemotional robot.
The eight Keepons in existence take turns to interact with toddlers and, with the help of scientists, react to their movements and exaggerated facial expressions that come at that age.
"This robot can recognize a face with a welcoming expression, for example," said Hideki Kozima, Keepon's creator and senior research scientist at Japan's National Institute of Information and Communications Technology.
Since 2003, Keepon has been used to study how children develop social behavior.
"This robot is meant to help child psychologists understand how children develop socially at the age of two or three," Kozima said.
"They start communicating with each other, recognizing facial expressions, and developing feelings of empathy. We are trying to incorporate that kind of social behavioral understanding in robots."
Keepon has two tiny cameras for eyes and a microphone for a nose. The squishy robot, hand-high and almost empty inside, is mounted atop a barrel which conceals four motors and two circuit boards, the main robotic parts.
Keepon won this year's Human-Robot Interaction Challenge at the IEEE International Conference for Robotics and Automation in Pasadena, California this month.
"There are a large number of papers that deal with human-robot interaction, a field in robotics that is generating a lot of excitement," said University of Southern California professor and conference co-chair Gaurav Sukhatme.
Marek Michalowski, a doctoral student at Carnegie Mellon University in Pennsylvania, has helped equip Keepon with complex features that make it less robotic.
"We want to make robots more human, or life-like," said Michalowski.
That sentiment is embraced by robotics engineers around the world and even by the entertainment industry.
Humanoid robots like Honda's ASIMO can perform complex tasks by imitating human-like gestures. Earlier this month ASIMO conducted the Detroit Symphony Orchestra with what bassist Larry Hutchinson described as "fluid" hand gestures.
Next month, Walt Disney Co. and Pixar Animation Studios will release the film "Wall*E" about a robot that looks nothing like a human, but which shows human-like emotions when it falls in love with a fellow robot named EVE.
The challenge for Kozima and Michalowski is not only to make Keepon more human, but also to make humans more comfortable with Keepon. In order to do that, they must equip Keepon with an awareness of non-verbal cues, like nodding or eyebrow raising.
"We often imitate the gestures of people we talk to," said Michalowski, as Keepon nodded in tandem with his repetitive hand gestures. "This is called rhythmic synchrony."
Michalowski and Kozima created a video of Keepon dancing to "I Turn My Camera On," a song by alternative-rock band Spoon that became a hit on You Tube (here).
So good was Keepon's groove that Spoon invited him to star in their video "Don't You Evah" as a dancer looking for a partner. (here).
(Editing by Mary Milliken and Patricia Reaney)
- Tweet this
- Share this
- Digg this