RealTime IT News

TV, Tell Me What's On

Web browser developer Opera Software wants to help you tell your TV to check what's on, thanks to the launch of a voice-enabled TV guide.

The Oslo, Norway, company's Electronic Program Guide (EPG) is based on IBM's ViaVoice implementation of the XHTML+Voice (X+V) multimodal markup language specification currently wending its way through the Worldwide Web Consortium (W3C) standards process.

The company plans to roll out a voice-enabled version of its desktop PC Web browser in late March or early April. Beta testing of the new browser began in December.

Officials said the technology is suited to any number of digital home entertainment components, such as DVD players and digital video recorders; it also focuses on the benefits voice technology would have on the cable or satellite set-top boxes. Now, instead of wading through programming on hundreds of channels by remote, users can tell the set-top box, "find me a documentary," or "find G-rated fare" in order to display the relevant shows.

Jon von Tetzchner, Opera CEO, said its Web-based presentation environment is ideal for EPGs because of the ease of customization, standards compliance and speed. Borrowing a page from Opera's customization abilities, operators and end-users can modify the EPG's themes and functionality with ease.

He said the integration of voice with data is a natural evolution, and has enormous potential in the integrated home media market. "Through our efforts with IBM, we hope to enable operators and OEMs to quickly get their HTML and Javascript-based applications talking."

Opera has proven itself adept at morphing its desktop Web browser for different media. Its desktop version, currently at version 7.54, has found its way into the mobile market through Qualcomm , Motorola and Microsoft as well as into embedded systems like set-top boxes.

Eskil Sivertsen, an Opera spokesman, said the company has a couple demonstration models of the voice-enabled EPG and is talking to manufacturers, but doesn't expect set-top boxes to feature its technology immediately.

"The whole media market is very complex and fragmented so it's not as simple as with desktop computers or mobile," he said. "It's a longer process."

Opera's software is based and developed on IBM's WebSphere Multimodal Toolkit, which uses the Everyplace Multimodal Environment to develop voice-enabled applications using the X+V markup language. X+V is made up of three W3C standards: XML Events, Voice XML and XHTML. Opera and IBM are the main authors of the X+V Profile version 1.0 specification.

Igor Jablokov, program director in IBM's multimodal and voice portals group, said the company's ViaVoice technology is slowly finding its way into different products, such as an announcement Monday with Teges Corp. to create an application that lets doctors and nurses enter or access patient information by voice or using a keyboard.

The next step, he said, is to get the technology in cell phones, which currently support a very rudimentary type of voice recognition. The goal isn't to get users to access pre-recorded information but to create a program that's ready to go from the moment its installed.

"I think we [need to] continue to enhance both our server-based technologies and our embedded-based technologies with enhanced features like user authentication," Jablokov said, "and things like natural language understanding, where you really don't have grammar that is pre-defined but you use the [V+X] model to figure out what you're asking for."

The news was announced as part of the SpeechTek conference that kicked off in San Francisco Monday.