Investigating recognition and preferences for self-voice on speech generating devices in autism

Abstract

A substantial number of children with autism cannot communicate by voice (Eigsti, de Marchena, Schuh, & Kelley, 2011; Kjelgaard & Tager-Flusberg, 2001). Voice enabled Augmentative and Alternative Communication (AAC) devices, also called Speech Generating Devices (SGDs) are often the only means available through which these children can communicate (Boesch, Wendt, Subramanian, Hsu, 2012); Goldsmith & LeBlanc, 2004; van der Meer & Rispoli, 2010). While a range of SGDs are currently available, most lack the vital ability to sound like the child using them (van Santen & Black, 2009). However, modern concatenative speech synthesis technology is being developed to mimic a given speaker's voice, by transplanting speech fragments from a recorded acoustic inventory and recombining these into speech-output 'using sophisticated algorithms' (Klabbers, Kain, & van Santen, 2010: el).