Peer-reviewed research
Montreal Forced Aligner: Trainable Text-Speech Alignment Using Kaldi
Michael McAuliffe et al., Proceedings of Interspeech 2017, 2017 · DOI 10.21437/Interspeech.2017-1386
Read the originalRead on October 17, 2026
- What it is
- A 2017 Interspeech paper that introduces the Montreal Forced Aligner, an open-source system that aligns a text transcript to recorded speech so that word and phone boundaries are timed. It evaluates the aligner against human-annotated boundaries on English conversational and laboratory speech.
- How much weight it carries
- Peer-reviewed conference paper with a DOI. It evaluates English speech only, and its accuracy claims are relative to two older open-source aligners, not to modern neural aligners.
Facts taken from this source
- The 2017 Interspeech paper introduces the Montreal Forced Aligner as a new open-source system for aligning speech with text.
“We present the Montreal Forced Aligner (MFA), a new open-source system for speech-text alignment.” (Abstract)
- The Montreal Forced Aligner was evaluated on word and phone boundaries against boundaries that humans had marked by hand.
“relative to human-annotated boundaries” (Abstract)
- The 2017 Montreal Forced Aligner paper says the aligner uses Kaldi instead of HTK, so that it can be distributed as a stand-alone package.
“MFA uses Kaldi instead of HTK, allowing MFA to be distributed as a stand-alone package” (Abstract)