arXiv Open Access 2022

Improving Automatic Speech Recognition for Non-Native English with Transfer Learning and Language Model Decoding

Peter Sullivan Toshiko Shibano Muhammad Abdul-Mageed

Lihat Sumber

Abstrak

ASR systems designed for native English (L1) usually underperform on non-native English (L2). To address this performance gap, \textbf{(i)} we extend our previous work to investigate fine-tuning of a pre-trained wav2vec 2.0 model \cite{baevski2020wav2vec,xu2021self} under a rich set of L1 and L2 training conditions. We further \textbf{(ii)} incorporate language model decoding in the ASR system, along with the fine-tuning method. Quantifying gains acquired from each of these two approaches separately and an error analysis allows us to identify different sources of improvement within our models. We find that while the large self-trained wav2vec 2.0 may be internalizing sufficient decoding knowledge for clean L1 speech \cite{xu2021self}, this does not hold for L2 speech and accounts for the utility of employing language model decoding on L2 data.

Topik & Kata Kunci

cs.CL cs.SD eess.AS

Penulis (3)

Peter Sullivan

Toshiko Shibano

Muhammad Abdul-Mageed

Format Sitasi

APA MLA BibTeX

Sullivan, P., Shibano, T., Abdul-Mageed, M. (2022). Improving Automatic Speech Recognition for Non-Native English with Transfer Learning and Language Model Decoding. https://arxiv.org/abs/2202.05209

Akses Cepat

Lihat di Sumber

Informasi Jurnal

Tahun Terbit: 2022
Bahasa: en
Sumber Database: arXiv
Akses: Open Access ✓