Classifier combination schemes in speech impediment therapy systems

Paczolay Dénes and Felföldi László and Kocsor András: Classifier combination schemes in speech impediment therapy systems. In: Acta cybernetica, (17) 2. pp. 385-399. (2005)

[thumbnail of Paczolay_2005_ActaCybernetica.pdf]
Preview
Cikk, tanulmány, mű
Paczolay_2005_ActaCybernetica.pdf

Download (216kB) | Preview

Abstract

In the therapy of the hearing impaired one of the key problems is how to deal with the lack of proper auditive feedback which impedes the development of intelligible speech. The effectiveness of the therapy relies heavily on accurate phoneme recognition [1, 4, 17]. Because of the environmental difficulties, simple recognition algorithms may have a weak classification performance, so various techniques such as normalization and classifier combination are applied to increase the recognition accuracy. This paper examines Vocal Tract Length Normalization techniques [5, 13] focusing mainly on the real-time parameter estimation [12], and the majority of classifier combination schemes, including the traditional (Prod, Sum, Min, Max) [7], basic linear (simple, weighted, AHP-based [6] averaging), and some special linear (Bagging, Boosting) combinations. Based on the results we conclude that hybrid combinations can improve the effectiveness of the real-time normalization methods.

Item Type: Article
Journal or Publication Title: Acta cybernetica
Date: 2005
Volume: 17
Number: 2
ISSN: 0324-721X
Page Range: pp. 385-399
Language: English
Place of Publication: Szeged
Event Title: Conference for PhD Students in Computer Science (4.) (2004) (Szeged)
Related URLs: http://acta.bibl.u-szeged.hu/38520/
Uncontrolled Keywords: Számítástechnika, Kibernetika
Additional Information: Bibliogr.: p. 398-399. ; összefoglalás angol nyelven
Subjects: 01. Natural sciences
01. Natural sciences > 01.02. Computer and information sciences
Date Deposited: 2016. Oct. 15. 12:25
Last Modified: 2022. Jun. 15. 12:41
URI: http://acta.bibl.u-szeged.hu/id/eprint/12772

Actions (login required)

View Item View Item