1

Context

I'm working with DragonFly, the python module. At this point, I just want speech recognized and outputted back to me.

Question

I'm wondering if there is a way to use DragonFly purely for voice recognition without going through Windows Speech Recognition (or any alternative program). Is Dragonfly only meant for "post-speech-recognition"?

The examples I've seen and run all open up windows speech recognition. I've also looked into the old speech recognition module - pySpeech, but that also "borrows" windows speech recognition.

Should I be looking towards other modules?

Luigi
  • 4,129
  • 6
  • 37
  • 57
sihrc
  • 2,728
  • 2
  • 22
  • 43

1 Answers1

3

DragonFly uses the shared recognizer, which starts Windows Speech Recognition. If you modify DragonFly to use an inproc recognizer, Windows Speech Recognition will not start. (Unfortunately, I don't know enough Python to contribute a fix.)

Eric Brown
  • 13,774
  • 7
  • 30
  • 71
  • Thanks for your reply. By inproc recognizer, do you mean a similar and adequate alternative to windows speech recognition? – sihrc Aug 26 '13 at 03:11
  • There are two ways to create a speech recognition engine. One is to create a "shared" recognizer, which has a number of defaults set, but which also starts Windows Speech Recognition; the other is to create an "inproc" (or in-process) recognizer, which is local to your process. However, the "inproc" recognizer doesn't have any defaults set (you have to specify the input source, recognition engine, etc.), as the assumption is that if you wanted the defaults, you would have created a shared recognizer. – Eric Brown Aug 26 '13 at 03:42