You are here  Software

IBM commits to speech technology

Tuesday 26 June 2001 01:56
IBM has renamed its portfolio of voice recognition systems Conversational Services.

The company will roll out products that will include speech translation, multimodal interfaces, middleware, natural-language understanding (NLU), text-to-speech and biometrics.

IBM will soon introduce one of the first products to use visual cues, such as lip and mouth movements, to understand the spoken word for speech interpretation, according to Dr David Nahamoo, senior manager of the human language technologies department at IBM's Thomas J Watson Research Center.

He added that the product is already in beta with a number of enterprises and would be available in about two years' time.

The visual recognition system could be of benefit in customer relationship management applications, for example, where call centre personnel would be able to interpret a customer's body language, Nahamoo said.

"The face is sending a message, happiness, sadness, anger. The challenge is how you model that and integrate it on top of the other [speech] technologies," he added.

In the short term, IBM's visual recognition system uses a microphone, a camera to monitor lip and mouth movement, and a set of business rules built into the recognition system.

"It might have a policy that if the face is not looking at the camera, the system understands that the person is not talking to me and so the computer can eliminate the sounds as noise," Nahamoo said. Also, if the lips are not moving but the system is picking up words or sounds, that information is filtered out as extraneous..

Some of these technologies will be especially useful in noisy environments, such as a moving car or on a stock market trading floor, noted Nigel Beck, IBM's director of Voice Systems.

"If the vocabulary in the system is small enough, it can recognise some words even in noise, and can be trained for digits in something as noisy as a 10-decibel environment," Beck said.

The system builds templates in time for each movement of the lips and converts the information into the basic ones and zeroes that computers understand.

The visual analysis is called a "viseme", not unlike a phoneme, the smallest intelligible segment of sound in a word. A viseme is the smallest intelligible segment of a lip gesture which, when put together with other visemes, allows the system to recognise the movements in aggregate as a word.

In other recent developments, last week IBM officials displayed a prototype add-on sled that will fit onto the back of a Palm handheld. The speech sled contains a DSP (digital signal processing) chip and memory for translating speech to text, and can be used for executing commands to a contact database or appointments calendar, as well as for voice-activated phone dialing.

IBM is also focusing efforts on its WebSphere middleware products. Last month it introduced its WebSphere Translation Server, which can translate about 500 words per second. An official said that Deutsche Bank will be one of the first customers.

Finally, IBM is using something called phrase slicing so that users will not have to listen to a mechanised voice.