Speech to Text

[sandyhardi]

Hi! We have a client that would like to be able to collect the "Voice Input" from caller.
For example, a caller calls in. Prompted, please say the department you would like to call: Support or Sales.
The Caller says, I want to talk to Sales.

The Client, wants us to get this Voice Input and transcribe it to text or collect the audio. is this possible?

 

[Evolute IT]

CFD.

 

[sandyhardi]

Yes, using CFD, but how to actually transcribe it?

 

[Evolute IT]

https://www.3cx.com/docs/manual/cfd-components/#h.lgzfai7rclby

 

[edossantos_sipcaller]

The Voice Input component lets you do the transcription of the spoken text, and also record it to a WAV file. This guide will help:
https://www.3cx.com/docs/cfd-google-cloud-speech/

 

[sandyhardi]

oh another thing; The Hints and Dictionary.

is the First line of a Hint corresponds to 1st line of Dictionary?2nd Hint entry corresponds and hints for 2nd entry in dictionary? so on and so forth? and can we put more separated by Comma? for example:

Hint 1: Yes, Yessss, Yez
Hint 2: No, Nouh, Nah


Dictionary 1: "Yes"
Dictionary 2: "No"

The Voice Input component lets you do the transcription of the spoken text, and also record it to a WAV file. This guide will help:
https://www.3cx.com/docs/cfd-google-cloud-speech/

 

[edossantos_sipcaller]

No, hints and dictionary entries are totally unrelated. The hints are just "suggestions" for the Speech Recognition engine, so in case of ambiguity it has information on what is expected, and then it can provide a better response. The dictionary entries are specific values that are expected, and 3CX will try to match the recognized text to any of them, and once it has a match it will return that value. So if the dictionary has the values "Yes" and "No", when the recognized text has any of those words, it will return the match.