IM this article to a friend!

September 24, 2004

New phone a sight for deaf eyes

From: Sydney Morning Herald, Australia - Sep 24, 2004

By Mike Barton
September 24, 2004

Lip-readers can now make virtual face-to-face calls with a telephone that mouths words uttered at the other end of the line.

A breakthrough for people with hearing difficulties, the SpeechView phone, made by an Israeli company, translates words as they are spoken into animation and displays them on a pop-up screen with a computer-animated face.

The $US200 ($280) phone is capable of being used in multiple languages as well as slang, although the first commercial version uses Hebrew. An English version will follow.

Also coming to the market is Synface, which can work with a phone plugged into a personal computer. It is a collaborative project funded by the European Union and trained to work in English, Swedish and Dutch.

A trial last month by Britain's Royal National Institute for Deaf People found that 84 per cent of participants could recognise the words mouthed by the Synface software's animated face.

By recognising sounds, speech-to-animation technology can generate lip movement animation in real time.

At the same time, a new standard for translating text into speech looks set to be a boon for visually impaired people. Called Speech Synthesis Markup Language, it allows information on the internet to be heard.

It would also provide "the next step beyond" the interactive voice response Telstra uses for directory assistance, the chief scientist for communications technology at the CSIRO, Trevor Bird, said.

Dr Bird said software that could connect the web to voice requests was relatively close.

The issue now, he said, was getting access to information that was stored on the internet and in private databases.

Last week, Timothy Berners-Lee, the inventor of the internet, told the SpeechTek conference in New York: "[The new standard] will allow an unprecedented number of people to use any telephone to interact with ... web-based services via keypads or spoken commands, and by listening to pre-recorded speech, synthetic speech and music."

Copyright © 2004. The Sydney Morning Herald.