Evaluation of a Speech Recognition Prototype for Speakers With Moderate and Severe Dysarthria: A Preliminary Report
By Fager, Susan K.; Beukelman, David R.; Jakobs, Tom; Hosom, John-Paul; Augmentative and Alternative Communication (AAC), Vol. 26, No. 4, pp. 267-277Publication Date: December 2010
Study describes preliminary work with the Supplemented Speech Recognition (SSR) system for speakers with dysarthria. The speaker dependent, discrete utterance SSR system developed for this work incorporated automatic speech recognition, alphabet supplementation, and word prediction. It was trained on the individual voices of the participants, using a single recording of each of 211 words. Recognition was performed using a total vocabulary of 500 words, including the 211 used in training. The software interface hosted by a tablet computer included an onscreen keyboard, message window, and word prediction buttons. Study participants were 7 speakers with dysarthria of varying severity aged 24 to 51 years, 6 of whom were male, with one participant with cerebral palsy demonstrating spastic dysarthria and the remaining 6 having mixed dysarthria following traumatic brain injury. Keystroke savings using SSR averaged 68.2 percent for typical sentences and 67.5 percent for atypical phrases. This was significantly different compared to using word prediction alone. The SSR correctly identified an average of 80.7 percent of target stimulus words for typical sentences and 82.8 percent for atypical phrases. Statistical significance could not be claimed for the relations between sentence intelligibility and keystroke savings or sentence intelligibility and system performance. Based on study results, the authors conclude that individuals with dysarthria using SSR could achieve comparable keystroke savings regardless of speech severity.
Published by: International Society for Augmentative and Alternative Communication (ISAAC) (Website:http://www.isaac-online.org)
