Replication Data for "The Mason-Alberta Phonetic Segmenter: A forced alignment system based on deep neural networks and interpolation"

These are the TextGrid outputs of the alignments run during the evaluation process for the MAPS paper. They are separated into train, validation, and test sets. The results from training and running the Montreal Forced Aligner (MFA) are also provided. The tar file for MFA contains subfolders for each of the train, validation, and test sets. The code that evaluated these files can be found on GitHub at https://github.com/MasonPhonLab/MAPS_Paper_Code.

If your operating system does not have an untarring utility pre-installed to open the tar files, 7-zip may prove a good option: https://www.7-zip.org/