Automatic Speech Evaluation and Multi-model Feedback Language Training to Articulation Disorders

Articulation errors will seriously reduce speech intelligibility and the ease of spoken communication. Typically, a speech-language pathologist uses his or her clinical experience to identify articulation error patterns, a time-consuming and expensive process. Moreover, the language training is very difficult to articulation disorders for personal training procedure. In this paper, a system with automatic speech evaluation and multi-model feedback language training is proposed to assist speech-language pathologists and articulation disorders. The articulation error patterns in phonetic can be identified and considered to generate pronunciation confusion network. Using speech recognition technique, the pronunciation errors can be automatically identified and labeled by dependency network. The articulation error pattern can be identified by dependency network. For language training, multi-model feedback interface is also proposed to assist articulation disorders. A 3D virtual facial animation with speech signal, lip motion, and tongue motion is proposed to promote users to articulate clearly. Moreover, the articulation teaching activities for each type of articulation errors is also feed back to caregivers. It can promote articulation disorders to improve articulatory abilities in games. Experimental results reveal the practicability of proposed method and system. Keyword: Articulation Disorder, Articulation Error Pattern, Multi-model Feedback, 3D Virtual facial animation, Dependency Network