Share via

  • Article

Note

Please see Azure Cognitive Services for Speech documentation for the latest supported speech solutions.

Microsoft Speech Platform

SPVALUETYPE

SPVALUETYPE lists flags indicating portions of a recognition result to be removed or eliminated once they are no longer needed.

<pre IsFakePre="true" xmlns="http://www.w3.org/1999/xhtml"> <strong>typedef enum tagSPDISCARDTYPES</strong> <strong>{</strong> <strong>SPDF_PROPERTY,</strong> <strong>SPDF_REPLACEMENT,</strong> <strong>SPDF_RULE,</strong> <strong>SPDF_DISPLAYTEXT,</strong> <strong>SPDF_LEXICALFORM,</strong> <strong>SPDF_PRONUNCIATION,</strong> <strong>SPDF_AUDIO,</strong> <strong>SPDF_ALTERNATES,</strong> <strong>SPDF_ALL</strong> <strong>} SPVALUETYPE;</strong> </pre>

Elements

  • SPDF_PROPERTY
    Removes the property tree.
  • SPDF_REPLACEMENT
    Removes the phrase replacement text for inverse text normalization.
  • SPDF_RULE
    Removes the non-top level rule tree information for a phrase.
  • SPDF_DISPLAYTEXT
    Removes the display form of the recognition result text.
  • SPDF_LEXICALFORM
    Removes the lexical form of the recognition result text.
  • SPDF_PRONUNCIATION
    Removes the pronunciation string of the recognition result.
  • SPDF_AUDIO
    Removes the audio data that is attached to a phrase. However, the audio must have been both set and retained.
  • SPDF_ALTERNATES
    Removes the alternate data that is attached to a phrase. Discarding recognition alternates loses the words permanently and they cannot be retrieved, even with ISpRecoResult::GetAlternates.
  • SPDF_ALL
    Removes all the elements above.