ADVERTISEMENT
See Inside June 2005

Conversational Computers

Efforts to make computers speak naturally will let machines better communicate meaning

Call a large company these days, and you will probably start by having a conversation with a computer. Until recently, such automated telephone speech systems could string together only prerecorded phrases. Think of the robotic-sounding "The number you have dialed ... 5 ... 5 ... 5 ... 1 ... 2 ... 1 ... 2...." Unfortunately, this stilted computer speech leaves people cold. And because these systems cannot stray from their canned phrases, their abilities are limited.

Computer-generated speech has improved during the past decade, becoming significantly more intelligible and easier to listen to. But researchers now face a more formidable challenge: making synthesized speech closer to that of real humans--by giving it the ability to modulate tone and expression, for example--so that it can better communicate meaning. This elusive goal requires a deep understanding of the components of speech and of the subtle effects of a person's volume, pitch, timing and emphasis. That is the aim of our research group at IBM and those of other U.S. companies, such as AT&T, Nuance, Cepstral and ScanSoft, as well as investigators at institutions including Carnegie Mellon University, the University of California at Los Angeles, the Massachusetts Institute of Technology and the Oregon Graduate Institute. Like earlier phrase-splicing approaches, the latest generation of speech technology--our version is code-named the IBM Natural Expressive Speech Synthesizer, or the NAXPRES Synthesizer--is based on recordings of human speakers and can respond in real time. The difference is that the new systems can say anything at all--including natural-sounding words the recorded speakers never said.

This is only a preview. Get the rest of this article now!

Select an option below:

Customer Sign In

*You must have purchased this issue or have a qualifying subscription to access this content


It has been identified that the institution you are trying to access this article from has institutional site license access to Scientific American on nature.com.
Click here to access this article in its entirety through site license access.

Share this Article:

Comments

You must sign in or register as a ScientificAmerican.com member to submit a comment.
Scientific American Holiday Sale

Limited Time Only!

Get 50% off Digital Gifts

Hurry sale ends 12/31 >

X

Email this Article

X