Built by Maven

TrsToKaldi

Converts Transcriber .trs files to corpus input files for Kaldi

The participant genders from the Transcriber transcripts are used, if present, to generate the spk2gender file. The following participant meta-data is lost during conversion:

  • dialect
  • accent
  • scope
  • version
  • version date
  • air date
  • scribe
  • language

The following Transcriber annotations are lost during conversion:

  • phrase language annotations
  • named entity annotations
  • comments
  • noises
  • lexical tags
  • pronounce tags

By default, all words are converted to lowercase, and extraneous punctuation is removed. To disable this behaviour, use the –cleanOrthography=false command line switch.

Deserializing from “Transcriber transcript” text/xml-transcriber

Command-line configuration parameters for deserialization:
--topicLayer=Layer Topic tags
--commentLayer=Layer Commentary
--noiseLayer=Layer Noise annotations
--languageLayer=Layer Inline language tags
--lexicalLayer=Layer Lexical tags
--pronounceLayer=Layer Manual pronunciation tags
--entityLayer=Layer Named entities
--scribeLayer=Layer Name of transcriber
--versionLayer=Layer Version of transcriber
--versionDateLayer=Layer Version date of transcriber
--programLayer=Layer Name of the program recorded
--airDateLayer=Layer Date the program aired
--transcriptLanguageLayer=Layer The language of the whole transcript
--participantCheckLayer=Layer Participant checked
--genderLayer=participant_gender Gender - participant ‘type’
--dialectLayer=Layer Participant's dialect
--accentLayer=Layer Participant's accent
--scopeLayer=Layer Participant's ‘scope’

Serializing to “Kaldi Files” text/x-kaldi-text

Command-line configuration parameters for serialization:
--orthographyLayer=orthography Orthography tags
--pronunciationLayer=Layer Pronunciation tags
--genderLayer=participant_gender Participant gender
--prefixUtteranceId=false Whether to prefix utterance IDs with the speaker ID or not.
--wavBasePath= Base path to prefix all wav files names.