Along with sharing your workload, soon your computers will also share your stress and emotions as Australian scientists are now working on developing voice recognition software capable of recognising emotions like stress, anger urgency etc, as conveyed by the user’s voice.

Current speech recognition technologies can sometimes detect when human intervention is needed, but need more finesse, says Professor Robert Dale, director of Macquarie University’s Centre for Language Technology.

“By monitoring aspects of the acoustic signal, a machine can make a guess at the emotional state of the caller,” says Dale, who predicts this will be standard within five years.But for now, some systems rely on recognising certain key phrases to tell when customers are fed up,” he was quoted by the ABC Online, as saying.

Psychologists, musicians, creative artists and linguists will join IT, science and engineering experts in a bid to unravel communication problems between humans, as well as those between humans and machines.

Scientists hope to find ways to improve the experience for those who get stuck talking to a machine. And chances are, we’ll have to do more machine talk in future, as more companies look for cheap ways to handle customer calls.

“One challenge is how to get a machine to realise it has made a mistake.One thing people tend to do is speak more slowly, but this only makes the problem worse. The machines have been trained for average speaking speeds,” Dale added.

More here.