Audio Processing of African Accented and Multilingual Speech
By Chibuzor Okocha on
Jul 31, 2025
This project adopts a broad, holistic view of African‐accented audio and speech processing, examining each stage and component that underpins robust, context‐aware voice technologies. By centering the rich diversity of African accents and linguistic patterns, it aims to advance both the scientific foundations and practical applications of modern Audio Language Models (ALMs). Some of the Audio task I am currently researching:
- Speaker Diarization
- Speech and Audio Quality using Audio Language Models
- African Accented ASR and Multilingual capacity of ALM
- Reasoning in Audio Language Models
- Application of ALM in domain specific use case like Engineering education
Demo
Forthcoming