I need to implement spoken command recognition in Android and iPhone apps without the use of internet communication - it all has to be processed locally on the devices and it must be able to distinguish 20-50 commands.
In other words: Local limited vocabulary speaker independent speech recognition running on both Android and iOS.
Using Google does not enlighten me enough so I was hoping for some advice from people having worked with this subject.
Any advice?
You can use CMUSphinx.
CMUSphinx on IPhone through OpenEars:
http://www.politepix.com/
CMUSphinx on Android:
http://cmusphinx.sourceforge.net/2011/05/building-pocketsphinx-on-android/