Microsoft Speech Recognition Achieves ‘Human Parity’
October 20th, 2016Via: Newsweek:
A “historic” landmark in artificial intelligence research has been reached by Microsoft, which has announced a new speech recognition technology that it claims is the first to be on par with humans.
Details of the software were published this week in a paper titled ‘Achieving Human Parity in Conversational Speech Recognition,’ authored by engineers at Microsoft Artificial Intelligence and Research.
“We’ve reached human parity,” said Xuedong Huang, Microsoft’s chief speech scientist. “This is a historic achievement.”
Harry Shum, who heads the Microsoft Artificial Intelligence and Research group, added: “Even five years ago, I wouldn’t have thought we could have achieved this. I just wouldn’t have thought it would be possible.”
According to the paper, the speech recognition system makes the same or fewer errors than professional transcriptionists. While it is the first time a computer has recorded a word error rate lower than a human, it is still far from perfect.
More work is needed to improve the system in real-life settings, such as places where there is a lot of background noise. Research into identifying individual speakers when multiple people are talking is also a part of longer-term research efforts.
