RAIL KNUST Participates in Deep Learning Indaba 2023

The Deep Learning Indaba is an annual African machine learning and AI community meeting to strengthen African AI. The Principal Investigator and Scientific Director, with three PhD scholars of the Responsible AI Lab (RAIL), participated in the Deep Learning Indaba 2023 in Accra at the CEDI Conference Center for a week-long learning, research, exchange and ideation event in machine learning and AI. The event featured keynote talks from ML/AI and Ethics experts, practical and hands-on programming sessions, mentorship, an African Research Symposium, and workshops on NLP, AI in Healthcare, Reinforcement Learning, Machine Learning, and AI Governance and Policy.

Amina Salifu, a PhD scholar of RAIL, presented on a transfer learning-based Ghanaian English accent recognition toolbox. She asserted the challenges in adequately transcribing or detecting Ghanaian English accents. She proposed using a transfer learning model to develop an efficiently and adequately transcribable Ghanaian English accent recognition toolbox. The technical presentation fostered collaborative learning and knowledge-sharing experience.

Prof. Jerry John Kponyo, PI and Scientific Director, RAIL, as part of the panel discussion on developing AI Policies for inclusive and responsible AI innovations, spoke on government policies for the responsible use of AI. He mentioned RAIL’s commitment to considering inclusivity and the ethical use of AI in its activities. 

“In RAIL, responsible AI means customer-centred AI,” he said. He mentioned the need for self-regulation as AI experts and researchers. 

During a keynote address on a quantitative approach to measuring responsible AI, Prof. Kponyo asserted that a quantitative measure limits subjectivity in evaluating AI solutions. “There is a need to provide a measure of how responsible an AI solution should be and an agreed standardised framework that helps compare AI solution performance globally,” he said. He introduced the FACETS framework as an attempt by the AI research labs in the AI4D initiative to propose a measure through which responsible AI could be measured, and it promotes and objectively measures the responsible nature of the activities and innovations from AI Labs.