Properties in MRCP/VXML/API

One difficult thing about developing applications using standards such as MRCP and VXML is that the names of similar parameters are often slightly different between the standards. This is particularly important when tuning voice activity detection parameters.

The following table presents a comparison of VAD properties in MRCP and their equivalent name in VXML and the LumenVox API (used with LV_SRE_StreamSetParameter and LVSpeechPort::StreamSetParameter).

In some cases the parameters do not have direct matches, so the exact mappings may differ depending on how a platform interprets the meaning of one specification or the other. Thus you should consider this table an informational guide rather than as an official interpretation of the standards.

  

MRCP

  
  

VXML

  
  

API

  
  

Sensitivity-Level 1

  
  

sensitivity 1

  
  

STREAM_PARM_VAD_VOLUME_SENSITIVITY

  
  

Speech-Complete-Timeout

  
  

completetimeout

  
  

STREAM_PARM_VAD_EOS_DELAY

  
  

Speech-Incomplete-Timeout 2

  
  

incompletetimeout 2

  
  

Partially supported by LumenVox 2

  
  

com.lumenvox.end-of-speech-timeout

  
  

N/A

  
  

STREAM_PARM_END_OF_SPEECH_TIMEOUT

  
  

No-Input-Timeout

  
  

timeout

  
  

STREAM_PARM_BARGE_IN_TIMEOUT

  
  

Confidence-Threshold

  
  

confidencelevel

  
  

N/A (handled by application)

  
  

Speed-Vs-Accuracy

  
  

speedvsaccuracy

  
  

N/A

  
  

N-Best-List-Length

  
  

maxnbest

  
  

PROP_EX_MAX_NBEST_RETURNED

  
  

Recognition-Timeout

  
  

maxspeechtimeout

  
  

N/A

  
  

Waveform-URL

  
  

N/A

  
  

N/A

  
  

DTMF-Interdigit-Timeout

  
  

interdigittimeout

  
  

N/A

  
  

DTMF-Term-Timeout

  
  

termtimeout

  
  

N/A

  
  

DTMF-Term-Char

  
  

termchar

  
  

N/A

  

Note that the sensitivity-level (sensitivity in VXML) works different with MRCP/VXML than in our API. In our API, setting this value higher makes it less sensitive to barge-in. MRCP and VXML are the opposite.

2 As of LumenVox versions 12.0, users can now elect to use the speech-incomplete-timeout value by enabling the 'use_speech_incomplete' option in the media_server.conf settings file.  If enabled, the greater of speech-complete-timeout and speech-incomplete-timeout will be used.


Was this article helpful?
Copyright (C) 2001-2024, Ai Software, LLC d/b/a LumenVox