Skip to content
QiTASC.com / Speech-to-Text Steps /
Introduction
/ .. /
Introduction










Speech to Text Steps

intaQt's speech-to-text functionality supports speech-to-text recognition of .wav audio files. This feature integrates services provided by the Google Cloud Platform Speech-to-Text API. In order to use the speech-to-text functionality via intaQt, Google Cloud Platform authentication must be set up, as described in Google's documentation.

Configuration

Syntax 

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
SpeechToText {
    enabled: <Boolean>
    gcpCredentialsFile: <String>
    profiles: {
        <ProfileName>: {
            language: <Language>
            phraseHints: [
                <List<phraseHints>
            ]
        }
    }
}

Parameters

  • enabled - true enables the Speech-to-Text functionality

    • The default is set to false

  • gcpCredentialsFile - A string representing either the absolute path or relative path to the project root of the Google Cloud Platform credential files

  • language (Optional) - The language to recognize, using a BCP-47 identifier

    • The default is set to en-US
    • The list of languages support by the Cloud Speech-to-Text platform is available at Language support

  • phraseHints - Contains a list of words and phrases that provide hints to the speech recognition task

    • Each phraseHint is limited to 100 characters per phrase and may not contain more than 500 phrases or 50,000 characters in total
    • Additional information is available at Content Limits

Example 

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
SpeechToText {
    enabled: true
    gcpCredentialsFile: "/opt/credentials/gcp.json"
    profiles: {
        speechProfile1: {
            phraseHints: [
                "welcome to your account"
            ]
        }
    }
}

Important! Failing to correctly set up the authentication will result in missing speech-to-text functionality.