ListSpeakers

Returns a list of speakers in a specified database. The action can also return information about the audio samples associated with each speaker.

Type: synchronous

Parameter Description Required
audiostatus A Boolean value (default false) that specifies whether to list the audio samples for each speaker. No
database The name of the speaker database to query. Yes
identifier A comma-separated list of identifiers to specify the speakers to list. If you do not set this parameter, Media Server lists all of the speakers in the specified database. No
metadata A Boolean value (default false) that specifies whether to return the metadata associated with each speaker. No

Example

The following example lists the speakers in the news database.

/action=ListSpeakers&database=news
                    &audiostatus=true
                    &metadata=true

Response

The following XML is an example response:

<autnresponse>
  <action>LISTSPEAKERS</action>
  <response>SUCCESS</response>
  <responsedata>
    <speaker>
        <identifier>newsreader</identifier>
        <numTrainAudio>1</numTrainAudio>
        <numDevelAudio>1</numDevelAudio>
        <metadata>
          <item>
            <key>category</key>
            <value>newsreaders</value>
          </item>
        </metadata>
        <audio>
          <label>665769c6de52a612b5de1ea4a8f1f1aa</label>
          <hasaudio>true</hasaudio>
          <training>true</training>
        </audio>
        <audio>
          <label>4a66aab4aace4ae39bbf1c5ffdaee33e</label>
          <hasaudio>true</hasaudio>
          <training>false</training>
        </audio>
        <modelBuilt>Yes</modelBuilt>
        <thresholdType>Automatic</thresholdType>
        <estimationBias>0.3</estimationBias>
        <threshold>34</threshold>
    </speaker>
    ...
  </responsedata>
</autnresponse>

The response can include the following information for each speaker:

  • identifier - The identifier for the speaker.
  • numTrainAudio - The number of audio samples that have been added to train the speaker model.
  • numDevelAudio - The number of audio samples that have been added to estimate speaker thresholds.
  • metadata - Custom metadata that you have added to the speaker using the action AddSpeakerMetadata. To see the metadata you must set the action parameter metadata to TRUE.
  • modelBuilt - Specifies whether the speaker model has been built. To build a model for a speaker, use the action BuildSpeaker. To build models for all speakers, use the action BuildAllSpeakers.
  • audio - Each audio element describes an audio sample that is associated with the speaker. This information is only present in the response when you set audiostatus=true.

    • The label element provides the label for the audio sample.
    • The hasaudio element indicates whether the audio sample is stored in the database. If this element contains the value false, the audio sample has been discarded and only the training is stored in the database. See NullSpeakerAudioData.
    • The training element indicates whether the audio sample was added for training (true) or for estimating speaker thresholds (false).
  • threshold (only present if the threshold has been set) - the threshold that is used to distinguish between a match to this speaker and an unknown speaker.
  • thresholdType (only present if the threshold has been set) - describes how the speaker threshold was calculated. If the threshold was set by running the action EstimateAllSpeakerThresholds or EstimateSpeakerThreshold, this element contains the value Automatic. If the threshold was set manually by running the action SetSpeakerThreshold, this element contains the value Manual.
  • estimationBias (only present if the threshold was set automatically) - the value used for the bias parameter when the speaker threshold was calculated with the action EstimateAllSpeakerThresholds or EstimateSpeakerThreshold.