Projekt Anti-Stotter

Abstract

The Anti-Stotter project demonstrated that AI-based stutter detection, fine-grained dysfluency segmentation, and automated anti-voice selection can be developed to a level of accuracy and efficiency compatible with real-time deployment on a wearable clinical device. The developed components have been handed over for integration into the FluxBuds® demonstrator and form the basis of the system's adaptive, self-learning behaviour.