About this Activity
This project involved fine-tuning multilingual self-supervised models for transcribing speech in Cameroonian languages, enabling the conversion of audio recordings into text. The goal was to create efficient tools for documenting and preserving local languages, which are often at risk of extinction.