A UKZN engineering graduate develops an AI prototype that recognises selected South African Sign Language signs and converts them into speech in real time. The system is designed to translate specific signs into spoken language as users communicate.
The reports focus on the prototype’s function—sign recognition and live speech output—rather than broader deployment. The available information does not specify the exact number of signs covered, the training data used, accuracy results, or whether the system runs on a dedicated device or standard hardware. There is also no detail on timelines for testing, scaling, or partnerships for wider rollout.
With only one outlet source provided, there are no differing angles across multiple reports to reconcile. The main emphasis across the coverage is the creation of an AI translation tool by a University of KwaZulu-Natal graduate and its intended accessibility benefit for South African Sign Language users.
