Add voice commands dynamically In unity

[wiwei]

Filing on behalf of another, reference #21759184

In order to use the voice command stuff in the MRTK, you have to define the actions and keywords up front (i.e. in the profile/editor). These then get fed to the dictation/speech recognizers (i.e. the array of words). There's a request here to be able to change these up at runtime, to handle the case where dynamically created buttons can also get voice commands.

[wiwei]

The tricky part about this one is the re-initialization of the speech providers needed to accomplish this (i.e. the set of keywords/voice commands are created at startup when new-ing up the keyword recognizer)

[Mt-Perazim]

is there any update on this feature?

[polar-kev]

This should be covered by #8310