Advancing African Language AI
Our research focuses on developing ethical, culturally-aware AI systems for African languages, with applications in social research, health, education, and financial inclusion.
Our Work
Current research and developments in African language AI, focused on real-world applications for social researchers
Fine-Tuning Automatic Speech Recognition Models for Wolof and Hausa in the Domain of Maternal and Reproductive Health
Authors: Ertony BASILWANGO, Oche ANKELI, Dhananjay BALAKRISHNAN, Yann LE BEUX
Venue: Deep Learning Indaba 2025
We present a fine-tuning approach for developing domain-specific automatic speech recognition models for Wolof and Hausa languages, with particular focus on maternal and reproductive health contexts. Our work demonstrates improved performance on health-specific tasks using small, curated datasets.
Research Datasets
High-quality, ethically sourced datasets for African language research and development
Social Research Audio Corpus (Wolof)
A curated dataset of Wolof audio recordings from health, education, and financial inclusion contexts
Collaborate with Us
We welcome collaborations with researchers, institutions, and communities working on African language technology. Together, we can build more inclusive and culturally-aware AI systems.