Journal of Information Technology & Software Engineering
Open Access

Speech systems to help persons suffering from a communication disorder

2^nd Global Summit and Expo Multimedia & Applications

August 15-16, 2016 London, UK

Kamil Kadi

University of Moncton, Canada

Posters & Accepted Abstracts: J Inform Tech Softw Eng

Abstract :

Communication is a multidimensional dynamic process that is necessary to express thoughts, emotions and needs. If one or more function involved in the communication process is impaired, the communication is disordered. A communication disorder has a large impact on the life quality; it reduces the capacity to express personality, exercise autonomy and often has an impact on relationships and self-esteem. Therefore, it is important to enhance the communication quality of individuals suffering from a verbal communication disability by offering them more possibilities to interact with their environment. Our research focuses on one of the most common speech communication disorders associated with a neurological impairment called dysarthria. In fact, millions of children and adults suffer from acquired or congenital neuro-motor communication disorders that can affect their speech intelligibility. The automatically characterization of speech impairment can contribute to improve the patient's life quality, and assist experts in assessment and treatment design. First, an original dysarthric speaker recognition system is proposed in a context where the individuals affected by verbal communication disorders are excluded from speech-enabled biometric solutions. Second, new approaches are implemented to improve the effectiveness of automatic diagnosis and assessment of the dysarthria severity levels. For this purpose, different levels of speech parameters were used, among them; computational models of auditory perceptual knowledge are proposed. Data are from two important worldclass databases of speech disorder, namely, Nemours and Torgo. Presented systems were developed using standard methods of machine learning and discriminant analysis. The results are promising for the proposed ASR system even though their preliminary nature; it can be used by patients to better communicate in a world where the importance of biometrics is growing. Also the algorithms of dysarthria assessment can be useful for therapists. The objective assessment, independent of the therapist may change the way of working with the patient.

Biography :

Email: kamil.kadi@NETAS.com.tr