Tradition or Innovation: A Comparison of Modern ASR Methods for Forced Alignment

TL;DR

This paper compares modern ASR methods WhisperX and MMS with traditional GMM-HMM-based MFA for forced alignment, revealing that traditional methods currently outperform modern end-to-end ASR systems on key datasets.

Contribution

It provides a direct performance comparison between leading modern ASR approaches and traditional GMM-HMM models for forced alignment tasks.

Findings

MFA outperforms WhisperX and MMS in alignment accuracy.

Modern ASR systems show limitations in forced alignment performance.

Traditional GMM-HMM models remain superior for FA tasks.

Abstract

Forced alignment (FA) plays a key role in speech research through the automatic time alignment of speech signals with corresponding text transcriptions. Despite the move towards end-to-end architectures for speech technology, FA is still dominantly achieved through a classic GMM-HMM acoustic model. This work directly compares alignment performance from leading automatic speech recognition (ASR) methods, WhisperX and Massively Multilingual Speech Recognition (MMS), against a Kaldi-based GMM-HMM system, the Montreal Forced Aligner (MFA). Performance was assessed on the manually aligned TIMIT and Buckeye datasets, with comparisons conducted only on words correctly recognized by WhisperX and MMS. The MFA outperformed both WhisperX and MMS, revealing a shortcoming of modern ASR systems. These findings highlight the need for advancements in forced alignment and emphasize the importance of integrating traditional expertise with modern innovation to foster progress. Index Terms: forced alignment, phoneme alignment, word alignment