Towards End-to-End Training of Automatic Speech Recognition for Nigerian Pidgin

TL;DR

The paper addresses end-to-end ASR for Nigerian Pidgin English, a high-need but under-resourced language, by creating a publicly available parallel speech-to-text dataset and evaluating three pretrained architectures. The key metric is word error rate (WER), defined as $WER = (S + D + I)/N$, and Wav2Vec XLSR-53 achieves the best test WER of $0.296$ after fine-tuning, outperforming Nemo QuartzNet and Wav2Vec 2.0 Base-100h. Zero-shot evaluation with an XLSR-English baseline yields $WER = 0.737$, which improves by approximately $59.84\%$ after adaptation to Nigerian Pidgin. The dataset and model weights are publicly released to support further research and to promote inclusive speech technology for under-resourced African languages.

Abstract

The prevalence of automatic speech recognition (ASR) systems in spoken language applications has increased significantly in recent years. Notably, many African languages lack sufficient linguistic resources to support the robustness of these systems. This paper focuses on the development of an end-to-end speech recognition system customized for Nigerian Pidgin English. We investigated and evaluated different pretrained state-of-the-art architectures on a new dataset. Our empirical results demonstrate a notable performance of the variant Wav2Vec2 XLSR-53 on our dataset, achieving a word error rate (WER) of 29.6% on the test set, surpassing other architectures such as NEMO QUARTZNET and Wav2Vec2.0 BASE-100H in quantitative assessments. Additionally, we demonstrate that pretrained state-of-the-art architectures do not work well out-of-the-box. We performed zero-shot evaluation using XLSR-English as the baseline, chosen for its similarity to Nigerian Pidgin. This yielded a higher WER of 73.7%. By adapting this architecture to nuances represented in our dataset, we reduce error by 59.84%. Our dataset comprises 4,288 recorded utterances from 10 native speakers, partitioned into training, validation, and test sets. This study underscores the potential for improving ASR systems for under-resourced languages like Nigerian Pidgin English, contributing to greater inclusion in speech technology applications. We publicly release our unique parallel dataset (speech-to-text) on Nigerian Pidgin, as well as the model weights on Hugging Face. Our code would be made available to foster future research from the community.

Figures (1)

• Figure 1: This shows the topic distribution in the Nigerian Pidgin text dataset, revealing dominant themes such as “General/Everyday conversations” (24.4%), “Government/Politics” (18.6%), and “Sports” (14.1%), with less-dominant themes in areas such as “Telecommunication” (1.3%) and “Agriculture (1%).”