September 11th, 2013, 04:28 AM
I actually have a new team member who did this in his last job. IIRC, he said he used a speech recognition library called Sphinx developed at Carnegie Mellon, for his application. You can google for it.
Bear in mind that it is not just the software though -- there needs to be a certain level of hardware support for speech recognition to work decently. For example, there's a very good *technical* reason why Siri is only for iPhone 4S and beyond and not the iPhone 4. It has to do with a noise reduction hardware chip, whose purpose to read only the user's voice and filter out all the secondary noise around the user. The iPhone 4 has an older version of this chip and the iPhone 4S and above have a newer version of this chip and they can read the user's voice without the user having to hold the phone to his/her face (the vendor for this chip is Audience Technologies).
Similarly, for Android, certain devices have better hardware than others and your speech recognition can only work if the hardware is "good enough" to support it.
In my colleague's case, he didn't need to worry about supporting a variety of different hardware devices, as the software was meant for a customer who standardized on one device and had a large budget (sorry, can't go into more details
Up the Irons
What Would Jimi Do? Smash amps. Burn guitar. Take the groupies home.
"Death Before Dishonour, my Friends!!" - Bruce D ickinson, Iron Maiden Aug 20, 2005 @ OzzFest
Down with Sharon Osbourne