Click here to return to the Bright Angles website

Thursday, 9 August 2012

Voice recognition technology - the new smartphone interface

Voice recognition technology has been around for a long time, but really took off with Apple's launch of the iPhone 4S and the debut of Siri, the iOS voice assistant. From being a novelty factor, Siri has progressed to being recognised as a useful and futuristic interface for smartphones - one of the first big steps taken towards making Artificial Intelligence accessible for daily use, and for the average user. 

Not surprisingly, Google has followed suit, debuting voice-based search with the newest Android OS, Jellybean. While inspired by Siri, Google's version is more oriented to their bread and butter search business. 

The challenges of voice recognition technology are manifold. The most basic barrier is recognition of the user's language and learning individual accents and pronunciation. At the next level, the software has to be intelligent enough to 'understand' what the user is really asking  rather than relying on fully formed questions. At the back end, servers have to provide sufficient power to process and answer queries in real time. And finally, the software needs an intelligent 'learning' component which builds on databases to return better and better answers to questions.

Voice recognition could play a huge role in India, where mass consumers are not always literate or even if they are, prefer talking rather than sending SMS. But we are a country dominated by feature phones and with potentially low mobile internet connectivity, though it is set to grow rapidly. 

Of course, the additional challenge in India is the multitude of vernacular languages, each with their own dialects, and spoken with a range of accents. Still, it should not pose a problem - if Siri can speak multiple languages, an Indian voice recognition assistant should be able to speak at least one language (Hindi) and deal with different accents.