London, June 11 : Japanese researchers have unveiled a humanoid robot that can understand three humans shouting at once.
Hiroshi Okuno at Kyoto University, and Kazuhiro Nakadai at the Honda Research Institute in Saitama call their robot Asimo.
The researchers attribute Asimo's mind-blowing ability to the new software they have designed, called HARK.
They have revealed that the software involves eight microphones to work out where each voice is coming from and isolate it from other sound sources.
It later works out how reliably it has extracted an individual voice, and then passes it onto speech-recognition software to decode, say the researchers.
Okuno insists that the HARK-based robot can go beyond normal human listening capabilities.
"It can listen to several things at once, and not just focus on a particular single sound source," New Scientist magazine quoted him as saying.
Although the scientist fraternity knows the ability to focus on a single voice among many is known as the "cocktail party effect", Okuno calls it the "Prince Shotoku Effect".
"According to Japanese legend, Prince Shotoku listened to 10 people's petitions at the same time," he says.
Both Okuno and Nakadai say that the HARK software can follow three persons speaking simultaneously with an accuracy of 70-80 per cent, when installed into Honda's Asimo robot.
The eight microphones around the Asimo's face and body enable it to accurately detect and isolate simultaneous voices.
"The number of sound sources and their directions are not given to the system in advance," says Nakadai.
The researcher, however, found that their software could identify only 30 to 40 per cent of what three persons shouting out restaurant orders in complicated sentences said.
The Asimo's ability is being used to judge rock-paper-scissors contests, where three people call out their choices at once.
Okuno and Nakadai presented their work at the 2008 IEEE International Conference on Robotics and Automation in Pasadena, California, last month.