The Montreal Forced Aligner can use Voice Activity Detection (VAD) capabilities from Kaldi to generate segments from a longer sound file.
Steps to create segments:
- Provided the steps in Installation have been completed and you are in the same Conda/virtual environment that MFA was installed in.
- Run the following command, substituting the arguments with your own paths:
mfa create_segments corpus_directory output_directory
The default configuration for VAD uses configuration values based on quiet speech. The algorithm is based on energy, so if your recordings are more noisy, you may need to adjust the configuration. See Create segments configuration for more information on changing these parameters.
Path to a YAML config file that will specify the alignment configuration. See Align configuration for more details.
Temporary directory root to use for aligning, default is
Number of jobs to use; defaults to 3, set higher if you have more processors available and would like to process faster