Today we were hosted by the Hong Kong Science Park in the New Territories for a fascinating tour of the cutting-edge research and development taking place there. An incubator for businesses involved in telecommunications, precision engineering and biotechnology, the science park is a brand-new research facility about a 45-minute drive north of the island. The CEO of the park gave us an overview of the project, then invited several representatives from three of their incubator projects to talk about their work.
Of the three presentations, one stood out in particular. Lawrence Mo of Kanhan technologies talked about his company's cutting-edge work in text-to-speech technology. Kanhan has developed a server-based technology for converting the text of websites written in English, Mandarin, Cantonese, Korean and Japanese into spoken language. Unlike VoiceXML technology, their system will work on any existing website, not requiring any retooling of existing websites.
Mo demonstrated a Hong Kong e-government portal that allows a users to hear a spoken version of the site in multiple languages. Rather than using client software, the website's text gets directed through Kanhan's server,which produces a streaming audio file of computer-generated spoken voice. This would allow a user to access spoken Web content on a computer that didn't have assistive screen reading software installed on it, such as a PC at a library or telecenter. The system is also able to "learn" new words that aren't in its pronunciation database, allowing its vocabulary to grow over time. The system also has a telephone gateway so Hong Kong people can call a local telephone number and listen to the text contained on the website.
The technology is part of a growing number of tools that have tremendous implications in bridging the digital divide. Creating intelligent tools for having website text read allowed and accessible over the phone opens up a whole new world to people with visual empairments, limited literacy and different language backgrounds.
I'm exploring the possibility of having the company experiment with my blog to use it as a test case for their text-to-speech technology. I'll let you know if it works out.... -ac
Posted by acarvin at May 13, 2004 07:48 AM | TrackBack