Speech recognition technology, now found mostly on desktop PCs, will soon branch out into more devices and products.
"More and more, we're driving technology into smaller devices," said Ozzie Osborne, general manager of IBM Voice Systems, at a press briefing last week. "This technology is getting so small that it can go anywhere."
At its research facility here, IBM previewed several prototypes of speech-recognition interfaces, including Personal Speech Assistant, a conversational user interface for Palm's Palm III personal digital assistant that includes IBM's Speech Coprocessor. Users can type or speak information and retrieve information by text, graphics or speech.
The company also demonstrated ViaVoice for Lotus Development's Notes, which responds to voice commands to compose, send, reply to and forward e-mail. Users can edit information in a calendar, an address book or a to-do list.
Company officials also previewed Speech Browser, a VXML (Voice Extensible Markup Language)-based device to search for information on company intranets or Web sites and conduct e-commerce on the Web.
The three products are due this year, officials said.
An IBM partner, California-based One Voice Technologies, demonstrated "human interface technology" that allows users to have a two-way dialogue with their PCs using natural language. Users can search the Web, launch desktop applications and access files.
The OEM and consumer versions of this product, which run on IBM's Via Voice Millennium speech engine, will be available this quarter.
Other speech recognition companies are also ramping up new products. Lernout & Hauspie Speech Products, will upgrade by mid year its Voice Express desktop speech recognition engine with improved accuracy. It also recently introduced a prototype handheld device with a large-vocabulary text-to-speech-recognition engine. The product, code-named Nak, will be ready within a year and cost US$500 or less, according to Bill DeStefanis, senior director of product management.
Dragon Systems is working on a prototype handheld device with embedded speech recognition. A company official said the product will be price-competitive when computer processing power is increased and memory access is improved.
Nuance Communications announced last week its Voice Web Portal Program to create voice-driven Internet content, commerce and services over the telephone.
IBM's speech-enabled prototypes
- Personal Speech Assistant: User interface for the Palm that includes IBM's Speech Coprocessor
- ViaVoice for Notes: Lets users compose, send, reply to and forward e-mail and edit a calendar, an address book or a to-do list
- Speech Browser: Web browser that supports VXML
