The microprocessor giant has been working on an advanced voice-recognition software platform for its x86- and XScale-based microprocessor lines. Within its labs in China, Intel is developing what it calls distributed speech-recognition (DSR) software technology, which supports both English and Chinese languages--particularly the Mandarin dialect (see April 13, 2001 story ).

Eventually, Intel hopes to develop more advanced speech-recognition technologies, such as universal translation machines. This application will require processors that run at speeds of 10-GHz or faster, according to the Santa Clara, Calif.-based chip giant.

And during the Intel Developer Forum (IDF) here this week, the company disclosed that it is developing other speech-recognition technologies, including what it calls "array-microphones" and "audio visual processing." "Array-microphones" promise to boost the audio quality in speech recognition, while "audio visual processing" is a lip-reading technology reminiscent of the sci-fi movie classic "2001: A Space Odyssey."

One of Intel's first commercial products in this arena will be a PC chip-set equipped with "array-microphone" technologies. Several vendors offer add-in cards with "array microphones," but Intel plans to integrate the technology within its core-logic lines, hinted Pat Gelsinger, senior vice president and chief technology officer for the company.

In 2004, Intel plans to offer a chip set with the technology, which hopes to solve a major problem with speech recognition applications in computers, Gelsinger said.

The problem with speech-recognition is that it's only 90-to-95% accurate. And it requires a controlled setting with little or noise in the environment, he said. "Speech recognition does well in low noise conditions," he said in an interview with SBN at IDF here this week.

Array microphones provide "highly directional signals," he said. "Array microphones also provide quality signals even in noisy environments," he said.

The proposed chip set, coupled with array-microphone technology, is expected to support current, third-party speech-recognition software offerings in the marketplace, according to Intel. It will also support Chinese, English, and other languages.

Intel declined to comment when this chip set would arrive in the marketplace, but the company did say that it is working on a longer-term project, based on "audio visual processing." In this project, Intel envisions a camera on a PC or system, which provides audio and visual "lip reading" technologies to boost voice recognition capabilities, Gelsinger said.

It's unclear how long before Intel commercializes the technology, which is reminiscent of the hit movie "2001: A Space Odyssey." However, in the movie, it must be remembered, the HAL computer on the spacecraft went awry after it felt threatened by ship's surviving crew member.

On the other hand, Intel's work in speech-recognition is important. The technology promises to solve some major problems, especially for Chinese-language populations in China, Hong Kong, Taiwan, and other regions.

For example, the Chinese language consists of over 10,000 characters. As a result, the Chinese-language keyboard on a PC is complex and cumbersome for users. Even the fastest typists in China can only type some 40 characters a minute on a Chinese keyboard, analysts said.

Speech-recognition could eradicate the need for the Chinese keyboard on a PC. While the jury is still out for speech-recognition products for the PC, the technology is showing signs of life in non-PC applications, like voice-portals, call-centers, and others.

Besides speech recognition, Intel is conducting research in other areas, especially wireless. At IDF, the company disclosed that it is pursuing research in meshed networks, smart antennas, software --defined radio, and ultrawideband.