Hello, It’s Your Phone Listening

Have you recently purchased a smartphone? If yes, have you tried talking to it? You heard it right, talking to it and not talking into it. Why, you ask? Well, because you can. Phones have been able to understand basic voice commands for a while now. But try using the latest generation of voice commands and you will be surprised at how good they are.

I tried using Google Voice Search on a relatively recent Android device. While looking up numbers for a Chinese takeaway from the Business Today office, the voice recognition software not only understood what I was saying –and believe me, I tried several different voices– it also used the geolocation abilities of the HTC Sensation that I was using to further filter results specific to my location.

But it is not just voice-based search that is revolutionary. Some of the best work in voice recognition is being done by Nuance, an American firm. Some of you must be familiar with Nuance’s Dragon Naturally Speaking software. Not only can this rather “amazing”tool help one input text without using a keyboard, the latest version also understands a whole host of accents and tones.

In fact, Nuance claims that the software constantly learns and adapts to the accent of the main user. Nuance has also developed voice transcription software which can understand medical terms. It is specifically aimed at the transcriptions and records market.

As a journalist, I find voice recognition tremendously useful. A one-hour interview often takes over five hours to transcribe, possibly longer, if one is dealing with a thick accent.

There is another very interesting piece of software which I recently tried on an Android device. The Vlingo Virtual Assistant, which works on all major smartphone operating systems, allows true handsfree driving with its ability to let users speak out a text message rather than type it, and also reads out messages. It is not perfect, but its search and speech read-out abilities make it quite useful. However, do keep in mind that using your mobile phone while driving, even with full handsfree ability, is dangerous and possibly illegal. That said, several luxury cars such as the Mercedes S-Class already come with certain controls that are commanded by the driver’s voice.

But the really interesting innovations in the voice recognition space are those in gaming consoles. Here products such as the Nintendo Wiimote and Microsoft Kinect, whose actions are read and understood by machines, can be combined with the voice control abilities of software. This has exciting prospects for the future, a future where the keyboard and the mouse become extinct and where the offices will be devoid of clickety-clack sounds.

Voice-based computing will have one pronounced impact in India, where we have a multitude of languages. Much as satellite television led to the spread of regional-language TV, local-language voice input on mobile phones or computers can lead to a dramatic rise in user adoption of the Internet. In China, all but a minuscule percentage of the Internet remains in English, mainly for expats. In India, only around five per cent of the Internet is in local languages. Voicebased inputs will allow people to create content in local languages far more easily than ever before.