Improve the speech recognition capabilities

[Rodrigo_L_]

Hey there, i was discovering more about the speech recognition feature, and was thinking, there's a way to improve the dictation? like the windows speech recognition improve its recognition as much u use it, there's a way to do it with realsense to? or it do automatically?

And i have a doubt too, u guys could explain with a example how to use nbest to get the other probably recognition?

I appreciate if u guys can help me.

[Xusheng_L_Intel · Accepted solution]

Speech recognition improvement works automatically. You may use ProfileInfo speaker field to create adaptation profiles for different users on the same machine. If you provide speaker name, the SDK module will adapt speech recognition to the speaker. Nbest is available for command and control mode only. You cannot use nbest in dictation mode. Example how to use nBest is in DF_Speech_recognition sample (speech_recognition.cpp, class MyHandler). Thanks!
 

[Rodrigo_L_]

Ok, thank you, i'll check about the profiles like u said, helped me a lot, thanks.