By Deborah Asantewaah SARFO
The University of Ghana (UG), in collaboration with Google Research Africa and UCL’s Global Disability Innovative Hub, is set to leverage cutting-edge AI technology to support persons with speech impairment to communicate easily.
The project, dubbed ‘tɛkyerɛma pa’ (‘good tongue’), is an AI-based initiative. It builds on the success of the Project Relate app, which continues to improve communication for individuals with non-standard speech worldwide.
It is a collaborative effort led by UG, supported by a US$40,000 grant from Google and research expertise from UCL’s GDI Hub provided through the AT2030 programme.
The announcement of the initiative was made known at a tech talk and press event held at the Google Ghana office.
In making communication inclusive for all irrespective of one’s level of literacy, it focuses on improving AI-powered speech recognition technology for five major Ghanaian languages – Akan, Ewe, Dagaari, Dagbani and Ikposo.
Globally, many individuals with non-standard speech due to conditions like cerebral palsy, ALS, stroke, down syndrome or Parkinson’s can face significant barriers to communication. Current speech recognition technologies often struggle to accurately interpret non-standard speech, limiting their effectiveness.
According to the Project Lead who doubles as the Associate Professor at UG, Professor Isaac Wiafe, the project will commence with data collection of non-standard speech in Akan. He added that they hope to use the data to train their models.
“We are starting to collect non-standard speech data in Akan. Our interest will be in the Asante language. We will be starting from Kumasi,” he said.
To make speech recognition accessible for all, Mr. Wiafe said the app is being designed to be run locally on mobile phones without the use of Internet.
The Senior Research Scientist at Google Research, Katrin Tomanek, emphasisied her team’s readiness to provide technical assistance and share insights learnt from their work in Project Euphonia to help local researchers create ASR models for local languages.
For his part, Prof. Cathy Holloway, Academic Director – GDI Hub, indicated that the impact of the collaboration will go beyond the country and further lay the “foundation for future language models and collections in Africa”.
Sharing her thoughts on the impact of the ‘tɛkyerɛma pa’ project, Adwoa Amponsah Dapaah, a lawyer with cerebral palsy, said localising the speech-recognition AI initiative will target all persons with speech impairment despite their educational background.
She added that the initiative will enforce inclusion, adding: “Inclusion does not mean only for persons with disability, but also someone with no educational background to communicate”.
The initiative is part of Google’s broader commitment to advancing AI development in Africa. By providing funding, technical expertise and cloud computing resources, Google aims to foster an ecosystem where local researchers can lead innovations that address both local and global challenges.
