Jump to: navigation, search

ASR Settings

If enabled for your account, the ASR Settings tab allows you to set the parameters for Automatic Speech Recognition by dragging the slider. If you make changes and want to return to the default values, click Reset.

ASR Settings Tab (Voicemail page example)

The following table describes those parameters:

Property Description
confidencelevel Specifies the speech recognition confidence level. Results are rejected (a nomatch event is thrown) when the caller's response is below this threshold. A value of 0.0 means minimum confidence is needed for recognition. A value of 1.0 requires maximum confidence.
sensitivity Specifies the sensitivity level. A value of 1.0 means that speech recognition is highly sensitive to quiet input. A value of 0.0 means that speech recognition is least sensitive to noise.
speedvsaccuracy Specifies the desired balance between speed and accuracy. A value of 0.0 means quick recognition. A value of 1.0 means best accuracy.
timeout Specifies the timeout in seconds after which a noinput event is thrown.
completetimeout Specifies the required length of silence (in seconds) following user speech before the recognizer finalizes a result (either accepting it or throwing a nomatch event).

Note: Values set for the completetimeout property are only supported if they are less than the values set for the incompletetimeout property.

incompletetimeout Specifies the required length of silence following user speech after the recognizer finalizes a result. The incompletetimeout property is used when the speech prior to the silence does not match all active grammars. In this case, once the timeout is triggered, the partial result is rejected with a nomatch event. The incompletetimeout property also applies when the speech prior to the silence matches an active grammar, but where it is possible to speak further and still match the grammar. By contrast, the completetimeout property is used when the speech prior to the silence matches an active grammar and no further words can be spoken.
maxspeechtimeout Specifies the maximum amount of speech input allowed (in seconds) after which an event is thrown leading to a nomatch.
Retrieved from "https://docs.genesys.com/Documentation:CXB:User:ASR:Current (2019-09-20 17:52:48)"
This page was last modified on July 29, 2016, at 03:33.

Feedback

Comment on this article:

blog comments powered by Disqus