Automatic Speech Recognition of African American English: Lexical and Contextual Effects

TL;DR

This paper investigates how African American English features, specifically CCR and ING-reduction, affect ASR accuracy, revealing that lexical effects are more prominent in systems without external language models.

Contribution

It provides new insights into the influence of specific AAE features on ASR performance and compares the effects with and without external language models.

Findings

CCR and ING-reduction significantly affect Word Error Rate

ASR systems without LMs are more influenced by lexical neighborhood effects

Lexical effects are small but statistically significant

Abstract

Automatic Speech Recognition (ASR) models often struggle with the phonetic, phonological, and morphosyntactic features found in African American English (AAE). This study focuses on two key AAE variables: Consonant Cluster Reduction (CCR) and ING-reduction. It examines whether the presence of CCR and ING-reduction increases ASR misrecognition. Subsequently, it investigates whether end-to-end ASR systems without an external Language Model (LM) are more influenced by lexical neighborhood effect and less by contextual predictability compared to systems with an LM. The Corpus of Regional African American Language (CORAAL) was transcribed using wav2vec 2.0 with and without an LM. CCR and ING-reduction were detected using the Montreal Forced Aligner (MFA) with pronunciation expansion. The analysis reveals a small but significant effect of CCR and ING on Word Error Rate (WER) and indicates a stronger presence of lexical neighborhood effect in ASR systems without LMs.