acm-header
Sign In

Communications of the ACM

ACM TechNews

Mimic-Bots Learn to Do As We Do


View as: Print Mobile App Share:

Korea Advanced Institute of Science and Technology (KAIST) researchers have developed a robotic system that responds to the actions of the person confronting it. The robot observes the person, analyzes the action, and stores it in its memory, all while making an appropriate response. The robot also develops a set of connections associated with a certain action, such as the action "throw" is connected to the object "ball." The robot also can request more gestures from the human if it is confused by conflicting information.

KAIST's Ji-Hyeong Han says the gesture-recognition technology overcomes the limitations of speech-based commands. "Of course, robots can recognize human intentions by understanding speech, but humans would have to make constant, explicit commands to the robot," Han says. "That would be pretty uncomfortable."

Socially intelligent robots that can communicate with humans through gesture and expression will need to develop a mental model of the person they are dealing with in order to understand their needs, says Bristol Robotics Laboratory's Chris Melhuish. The KAIST researchers plan to test the system using a robot with sensors that can detect users' movements and gestures.

From New Scientist
View Full Article – May Require Free Registration

 

Abstracts Copyright © 2011 Information Inc., Bethesda, Maryland, USA


 

No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account