e-ISSN:0976-5166
p-ISSN:2231-3850


INDIAN JOURNAL OF COMPUTER SCIENCE AND ENGINEERING

Call for Papers

Aug 2019 - Volume 10, Issue 4
Deadline: 15 Jul 2019
Notification: 15 Aug 2019
Publication: 31 Aug 2019

Oct 2019 - Volume 10, Issue 5
Deadline: 15 Sep 2019
Notification: 15 Oct 2019
Publication: 30 Oct 2019

Indexed in

IJCSE Indexed in Scopus

ABSTRACT

Title : ISOLATED SPEECH COMMANDS RECOGNITION BY USING HYBRID APPROACH
Authors : Tomas Rasymas, Vytautas Rudžionis
Keywords : Hybrid speech recognition; recognizers combination; intelligent systems; classifiers evaluation.
Issue Date : Dec 2017-Jan 2018
Abstract :
This paper presents our results obtained by experimenting with different classifiers and different types of features to combine different speech recognizers in order to increase isolated speech commands recognition accuracy. Two different speech corpora were used to evaluate classifiers and features performance: “Asmens kodai” – person code which consists of 11 digits, “Skaiciai” – digits from 0 to 9. Compared suggested speech recognizers combination method with best single recognizer accuracy increased based on corpora: “Asmens kodai” – 0.352 % and “Skaiciai” – 1.057 %. Compared with combination that was made using voting method accuracy increased based on corpora: “Asmens kodai” – 18.7 % and “Skaiciai” – 2.9 %. Classifiers that were evaluated in different speech recognizers combination task: Naďve Bayes, Random forest, Nearest neighbors, CART and Support vector classifiers. Highest accuracy increase obtained by using Random forest classifier.
Page(s) : 679-685
ISSN : 0976-5166
Source : Vol. 8, No.6