The NaijaVoices Dataset: Cultivating Large-Scale, High-Quality, Culturally-Rich Speech Data for African Languages

TL;DR

The NaijaVoices dataset provides a large-scale, diverse speech corpus for African languages, enabling significant improvements in speech recognition models and addressing the under-representation of these languages in technology.

Contribution

We introduce the NaijaVoices dataset, a large, high-quality speech-text resource for African languages, with a novel data collection approach and demonstrated impact on speech recognition performance.

Findings

Achieved up to 75.86% WER reduction with Whisper model

Demonstrated dataset's acoustic diversity and scalability

Enhanced multilingual speech processing for African languages

Abstract

The development of high-performing, robust, and reliable speech technologies depends on large, high-quality datasets. However, African languages -- including our focus, Igbo, Hausa, and Yoruba -- remain under-represented due to insufficient data. Popular voice-enabled technologies do not support any of the 2000+ African languages, limiting accessibility for circa one billion people. While previous dataset efforts exist for the target languages, they lack the scale and diversity needed for robust speech models. To bridge this gap, we introduce the NaijaVoices dataset, a 1,800-hour speech-text dataset with 5,000+ speakers. We outline our unique data collection approach, analyze its acoustic diversity, and demonstrate its impact through finetuning experiments on automatic speech recognition, averagely achieving 75.86% (Whisper), 52.06% (MMS), and 42.33% (XLSR) WER improvements. These results highlight NaijaVoices' potential to advance multilingual speech processing for African languages.