The World Wide Consortium or should I say W3C is thinking of giving soul to the WebPages in future. It does sound a little devilish right? Think of it… how does the idea of interacting with web pages that can listen to you, feel?
The W3C has its bunch of nerds with horn rimmed glasses, all teamed up to develop voice recognition based WebPages. Exactly one year from now, the group will submit its findings regarding the feasibility of adding voice and speech features into the W3C standards for rendering WebPages online.
This way, if the plan is supposed to work out, Google, AT&T, Microsoft, and Yahoo will have their engineers on their toes. Google already did a little bit of voice recognition gimmickry before and we can find hints of it in the Android smartphone market as well. Some evidence also suggests that Microsoft will be adding voice recognition support to a very limited aspect of Windows 7 Mobile OS.
But Dan Burnett – Director of the Speech Recognition Group at W3C is thinking of something else. He believes that a proper implementation of the technology can do wonders across multiple web browsers. He gave the idea of using a built in or a plugin based speech recognition software so that browsers could read aloud pages or permit users to fill in the relevant field through their voice.
While there are strong probabilities of overlapping the VoiceXML W3C effort, the group is willing to make the jump. The W3C is shortly going to release a new version of the Speech Synthesis markup Language often used in conjunction with the VoiceXML.
For browser-based voice recognition to be effective (i.e. accurate) it will need to support 2 features:
1) identify the speaker, and where their audio profile (in an open format) is located.
2) identify the content, in a way that supports identifying any custom grammar, lexicon, or methods (like using Google search term frequency info).
Both features improve recognition dramatically. They were anticipated (and patented) by MeMeMe’s chief scientist 10 years ago, and are available today using MeMeMe’s on-demand speech recognition service — voice-enabling any application, for any user, on any connected device.
Peter
CEO, MeMeMe