MFA

The MFA annotator integrates with the Montreal Forced Aligner, which can use words with phonemic transcriptions, and the corresponding audio, to force-align words and phones; i.e. determine the start and end time of each speech sound within each word, and thus the start/end times of the words.

This API provides access to information about the currently installed version of the montreal-forced-aligner conda package, and also provides access to logs.

The following GET http enpoints are available:

mfaVersion
Provide the version number of the montreal-forced-aligner package currently installed. The response is plain text; the raw output from running the command: mfa version
validDictionaryNames
List valid pronunciation dictionary names. The response is a JSON-encoded array; the parsed output from running the command: model download dictionary
validAcousticModels
List valid acoustic model names. The response is a JSON-encoded array; the parsed output from running the command: model download acoustic
listLogs
List logs from forced-alignment runs. The response is a JSON-encoded array of log file names.
downloadLog?log
Downloads the given log file.
deleteLog?log
Deletes the given log file.
mappingToCsv?id
Provides access to the mapping between the given two layers, as a CSV stream. The query string should be a mapping ID returned by listMappings. The response is a CSV file containing all edit steps mapping all source annotations to target annotations, with the following fields:
deleteMapping?id
Deletes all data associated with the given mapping. The query string should be a mapping ID returned by listMappings.