automate-alignment

This project enables:

1. The generation of .lab files from .docx files whose paragraphs match the audio recordings to be used for forced alignment on the Montreal Forced Aligner (MFA).
2. The generation of a G2P dictionary for words not stored in the original dictionary.
3. The merging of multiple G2P dictionaries to obtain a broader one.

I've worked on this project as a Research Assistant under Dr Fatima Hamlaoui at the University of Toronto from June to August 2018.