By:
Nana Appiah Acquaye
The
Responsible AI Lab (RAIL) has showcased its work on inclusive speech
recognition for low-resource languages at the AI for Good Summit, highlighting
efforts to make artificial intelligence technologies more accessible and
representative for multilingual communities.
The
project, developed with support from the Geneva Graduate Institute, focuses on
improving speech recognition systems to better understand diverse speech
patterns and languages that are often underrepresented in mainstream artificial
intelligence applications.
Led
by Prof. Jerry John Kponyo, Principal Investigator of the Responsible
Artificial Intelligence Lab, the initiative seeks to address challenges faced
by communities where people communicate across multiple languages, including
environments such as Ghana where speakers frequently switch between local
languages and English during everyday conversations.
Prof.
Kponyo and his team, comprising Ama Branoa Banful, Juliet Arthur and Kenneth
Dotse, are working to develop AI systems that reflect real-world communication
patterns and improve access to speech-based technologies for diverse
populations.
RAIL
said the project aligns with its commitment to advancing responsible artificial
intelligence by ensuring that AI solutions are designed with inclusion,
accessibility and local realities in mind.
