Overview
ModelTalker is a speech synthesis software package designed to benefit individuals who are losing or have already lost their ability to speak. Developed by the Nemours Speech Research Laboratory, it allows users of Speech Generating Devices (SGDs) to communicate using a unique, personalized synthetic voice that closely represents their own. The system involves voice banking, where users record a set of standardized phrases and words. These recordings are then used to train an AI model to generate a synthetic voice. ModelTalker offers different generations of voice technology, including Gen3 voices that leverage advanced AI for enhanced expressiveness and compatibility across various platforms. Use cases include enabling individuals with ALS, MND, or other speech-impairing conditions to maintain their vocal identity and communicate effectively. ModelTalker supports integration with iOS, Android, and Windows-based SGDs, tablets, and phones, expanding its accessibility and usability.