Specifies the confidence threshold for recognition results computed while the selective_barge_in mode is active.
|
Value |
Integer: 0–999. DEFAULT: 500 |
|
How to set |
Shown in precedence order (see Rules of parameter precedence): In the browser, using the Vendor-Specific-Parameters (MRCPv1) or Hotword-Confidence-Threshold (MRCPv2) header of a RECOGNIZE request. In a parameter grammar. In Management Station set on the Nuance recognition service. If not using Management Station, set in a Recognizer configuration file (User-nrsxx.xml). |
|
Usage |
Seldom used because most applications do not require the feature. But when the feature is needed, this parameter is occasionally used for tuning the threshold. |
|
Performance |
Higher values ensure more accuracy at the cost of sometimes rejecting a valid key word spoken by a caller. Lower values avoid the erroneous rejections at the cost of sometimes accepting invalid key words. |
Recognizer assigns confidence scores to recognition results. If the score for a selective_barge_in recognition is less than the value of this parameter, the audio sample is rejected, and Recognizer continues to compute subsequent samples.
The confidence threshold is optional and if set in the following way (this is an MRCPv2 example):
Confidence-Threshold:0.5
Recognizer interprets the setting as the swirec_selective_barge_in_conf_thresh parameter.
Related parameters
For general information on confidence scores, see confidencelevel.
