Leen Remmelzwaal will present the School of IT seminar with a  talk entitled, "A South African Case Study: Applying Machine Learning to Automatic Number Plate Recognition (ANPR).

Abstract: In this talk I will share how a South African company solved an expensive real-world problem with a light-weight mobile application: Automatic Number Plate Recognition (ANPR) for Law Enforcement. I will briefly discuss the ANN and CNN architectures, as well as the models and training techniques used to achieve an elegant mobile solution.

Profile: Leen Remmelzwaal is currently completing a PhD in Computational Neuroscience at UCT, exploring and modelling how the human brain is capable of intelligence and consciousness. His academic research and personal interests are in Machine Learning - building computers to think as well as humans do. Leen completed an MSc in Electrical Engineering and a BA Honours in International Relations. He is also a machine learning consultant with experience in supervised learning and image processing. Leen has 10+ years enterprise-level software engineering experience (including project planning, commercial budgeting, execution, deployment, team leadership, support, sales, contractual and legal).

Meeting link (MS Teams): https://teams.microsoft.com/l/meetup-join/19%3a2e1c0937c8334d1d98afe0fc95ad31fd%40thread.tacv2/1603652875617?context=%7b%22Tid%22%3a%2292454335-564e-4ccf-b0b0-24445b8c03f7%22%2c%22Oid%22%3a%2288b58923-2f32-4dde-adcd-61ad99fcd885%22%7d