Nigeria English Accent Audio Data Collection
Curated and processed audio datasets for Nigerian English accent, supporting ASR model fine-tuning for improved speech recognition accuracy.
Nigerian English is one of the most widely spoken English varieties in Africa, yet it is underrepresented in ASR training data. This project collects, curates, and processes audio samples of Nigerian English speech, with accurate transcriptions. The dataset is used to fine-tune models like Whisper and Wav2Vec2, significantly improving recognition accuracy for Nigerian speakers.