Enhancing Silent Speech Decoding Using AI for Assistive Communication Technologies

Silent speech interfaces (SSIs) transform inaudible gestures into written or spoken text, making SSIs beneficial for individuals unable to speak or those in silent spaces. Following the recent innovations in artificial intelligence (AI), particularly in deep learning, accuracy has increased; however, there are still gaps in the model’s robustness, real-time processing, and interpretability. This study aims to enhance SSIs by developing sophisticated AI algorithms to bolster accuracy, reliability, and privacy. The research helps integrate SSIs into practical assistive communication technologies for people with speech disabilities and others for broader scope of communication in silent environments.