Speech Quality
- 19 Jul 2024
- 1 Minute to read
- Contributors
- Print
- PDF
Speech Quality
- Updated on 19 Jul 2024
- 1 Minute to read
- Contributors
- Print
- PDF
Article summary
Did you find this summary helpful?
Thank you for your feedback
Measures the quality of the audio recording and assigns scores.
Parameters:
name | description | default |
|---|---|---|
| Address | Address of the speech quality service | http://core-speech-critique |
| IgnoreSslErrors | Ignore any certificate errors if emotion address contains https | false |
Inputs
Audio:
Accepts audio from a single channel.
Events:
none
Outputs
Audio:
none
Events:
| name | description |
|---|---|
| Speech Quality | Raised several times when the node has accumulated enough audio data to update speech quality measurement. The quality value is in range from 1 to 5. 5 means the highest quality. |
Remarks :
- This node evaluates the speech in segments of 3 seconds. Silent periods are not included in this duration.
Project Structure
The Speech Quality Node only needs audio input. The audio doesn't have to be segmented. So, you can place the Speech Quality node wherever the desired audio is flowing through. A simple project can be built as such:
However, a better approach would be:
Why this approach is better?
In the second example, the only difference is that we pass the audio through a VAD node. This means the audio is filtered. The only audio info that passes through a VAD is speech info. This approach will result in better confidence levels in Speech Quality identification since random silences are cleared.
Supported flow types: Stream, Batch
Was this article helpful?