Sunday, May 4, 2008

Will voice replace text?

There’s a new trend coming from software developers since CTIA Wireless conference to Web 2.0 Expo that were held not so long ago, that trend is the voice recognition service. Using what is the main function of our mobile phone; voice/talk & listen to it, these software developers see new ways to support it to even more advance than some basic usage of mobile phones.

Still, the purpose of using these new voice recognition features that are offered is only to lessen physical interaction of our fingers with mobile phone’s keyboard or its navigational pads. While we can’t stop looking at the screen yet, because it’s still the crucial need even in those voice recognition features.

Like what
VoiceCloud has to offer in its new voice-to-text service, which the VoiceCloud’s CEO; Gerald Marolda has boasted to Jessice Dolcourt (CNet) about his company’s fast and accurate service. The service is basically almost similar to Jott’s; but VoiceCloud has broader range of services that include voice mail, SMS & email. Marolda emphasis on VoiceCloud’s ‘human-touch’ services that distinguish them from other competitors, he quickly answered what’s the reason when asked on why he preferred people power. "Many companies claim to use voice-to-text software, but there is no technology in the market that exists right now that gives you the accuracy you need." VoiceCloud is also keen to jump fast enough into iPhone’s popularity band wagon, the free beta service offering has been optimized for iPhone; hit this link to get it.

The VoiceCloud service is indeed still requires the recipients on the other end, to look at their mobile phone’s screen when they received its pushed-text services. This has proven how essential it is the visual part in mobile phone usage, another example is
TellMe’s voice-to-command service which I’ve already told you about before here at PA. TellMe voice service technology is really based on its search ability over the air (OTA) to the main servers, but such service can be consider as a break through on minimizing the interactions on handsets & using your mobile phone only with your voice.

Voice recognition also has reached to GPS unit, where IMHO this is more needed because if you’re on the road looking for destination places; you don’t have to stop the car to input the name of the places or the roads anymore. Just shout out where you want to go, and the GPS unit will look it up by itself without taking your attention off the road. Again, voice recognition accuracy is what has been holding back GPS users to purchase such unit with feature like that. But Garmin’s nuvi 880, that I talked about before
here at PA too, has show a lot of potential to lead they way for other GPS units.

There are other similar voice-to-text services offered by many software developers such as SpinVox, SimulScribe, CallWave and Dial2Do. But these voice services lack of what’s the most important thing in voice technology: accuracy. Like Dial2Do that has just been reviewed by
MyTreo.net and posted here at PA by Sammy, Alli Flowers is sticking with Jott after hard times using its voice services.

Sprint also has tried with Voice Command service, which basically it’s a web-based address book that provides an alternative means for placing voice calls to contacts. But Sprint’s Voice Command was not easy to use at all, it would be more faster if we simply use traditional way to make calls instead of make a call to Voice Command centre first to place a call. According to
BGR, due to lack of customers interest on the service offered; Sprint is pulling the plug of Voice Command service on June 1st. BGR made a good advice about this: “Snag a modern phone with integrated voice recognition”.

Too bad right now, I haven’t heard of a worthy mobile phone that has superior embedded voice recognition feature. But in desktop OS, it’s not something new and in fact has come bundled in. I know, and must you; that Windows Vista has it. Not long ago,
jkOnTheRun has reported about new macro support for Vista Speech Recognition; “The idea is that you can script one or several tasks and essentially assign a voice tag to the macro”.

But that and with other speech feature updates for Vista, so far still doesn’t convince me entirely on Vista capability to process every speech given to it. The
Windows Vista Speech Recognition demo gone awry video has been glued to my mind ever since I first saw it, hit the link and you’ll know what I mean.

Voice recognition is definitely the future technology for modern mobile phone, it’s the newest booming trend in every Web 2.0 software developers. And they all are still offering their services through web, data transfer & require you to call to their call centre to use them. None, are offering it as embedded feature bundled inside new purchased mobile phone. I do believe that Microsoft will include Vista’s Speech Recognition technology into its new Windows Mobile 7, which dubbed as “iPhone-killer”.

If that’s to be happened, there is a good chance that we’ll be seeing less and less mobile phones with numeric pads or keyboard; the reason is pretty simple: voice recognition will help you do text messaging. And one thing left needed is larger touch-screen to display it; like what we have already seen as another trend in mobile phone by handsets such as HTC Touch, iPhone, LG Prada, and the upcoming monstrous-display
HTC Diamond.

Will Palm Nova support this voice recognition feature in its release? A bundled voice application is not necessary, although it would be a nice bonus, but OS platform that can support it is a must to have. This is not the only thing Palm should think long and hard before Palm Nova is released with new smartphone along with it, Palm should also think if it’s necessary to cut-off the Palm smartphone’s signature: the QWERTY keyboard.

Which one do you prefer more: a full QWERTY keyboard with standard touch-display (320x320) or visual keyboard on a large touch-screen? Or, in this current subject: voice recognition feature or text messaging?

[blogged with my Treo 750v]

No comments: