Robust digit recognition with dynamic time warping and recursive least squares

Robustness is a key issue in speech recognition. This paper proposes a speech recognition algorithm for Malay digits from 0 to 9. This paper also proposes an algorithm for noise cancellation by using recursive least squares (RLS). This system consists of speech processing inclusive of digit margin a...

Full description

Bibliographic Details
Main Authors: Al-Haddad, Syed Abdul Rahman, Ishak, Khairul Anuar, Abdul Samad, Salina, Abid, Ali O., Hussain, Noor Aini
Format: Conference or Workshop Item
Published: 2008
Online Access:http://psasir.upm.edu.my/id/eprint/16587/
_version_ 1848843003315093504
author Al-Haddad, Syed Abdul Rahman
Ishak, Khairul Anuar
Abdul Samad, Salina
Abid, Ali O.
Hussain, Noor Aini
author_facet Al-Haddad, Syed Abdul Rahman
Ishak, Khairul Anuar
Abdul Samad, Salina
Abid, Ali O.
Hussain, Noor Aini
author_sort Al-Haddad, Syed Abdul Rahman
building UPM Institutional Repository
collection Online Access
description Robustness is a key issue in speech recognition. This paper proposes a speech recognition algorithm for Malay digits from 0 to 9. This paper also proposes an algorithm for noise cancellation by using recursive least squares (RLS). This system consists of speech processing inclusive of digit margin and recognition which uses zero crossing and energy calculations. Mel-Frequency Cepstral Coefficient (MFCC) vectors are used to provide an estimate of the vocal tract filter. Meanwhile dynamic time warping (DTW) is used to detect the nearest recorded voice with appropriate global constraint. The global constraint is used to set a valid search region because the variation of the speech rate of the speaker is considered to be limited in a reasonable range, which means that it can prune the unreasonable search space. The algorithm is tested on speech samples that are recorded as a part of a Malay corpus. The results show that the algorithm can recognize almost 80.5% of the Malay digits for all recorded words. By adding RLS noise canceller in the preprocessing stage it increases the accuracy to 92.3%.
first_indexed 2025-11-15T08:08:07Z
format Conference or Workshop Item
id upm-16587
institution Universiti Putra Malaysia
institution_category Local University
last_indexed 2025-11-15T08:08:07Z
publishDate 2008
recordtype eprints
repository_type Digital Repository
spelling upm-165872016-02-03T07:05:47Z http://psasir.upm.edu.my/id/eprint/16587/ Robust digit recognition with dynamic time warping and recursive least squares Al-Haddad, Syed Abdul Rahman Ishak, Khairul Anuar Abdul Samad, Salina Abid, Ali O. Hussain, Noor Aini Robustness is a key issue in speech recognition. This paper proposes a speech recognition algorithm for Malay digits from 0 to 9. This paper also proposes an algorithm for noise cancellation by using recursive least squares (RLS). This system consists of speech processing inclusive of digit margin and recognition which uses zero crossing and energy calculations. Mel-Frequency Cepstral Coefficient (MFCC) vectors are used to provide an estimate of the vocal tract filter. Meanwhile dynamic time warping (DTW) is used to detect the nearest recorded voice with appropriate global constraint. The global constraint is used to set a valid search region because the variation of the speech rate of the speaker is considered to be limited in a reasonable range, which means that it can prune the unreasonable search space. The algorithm is tested on speech samples that are recorded as a part of a Malay corpus. The results show that the algorithm can recognize almost 80.5% of the Malay digits for all recorded words. By adding RLS noise canceller in the preprocessing stage it increases the accuracy to 92.3%. 2008 Conference or Workshop Item PeerReviewed Al-Haddad, Syed Abdul Rahman and Ishak, Khairul Anuar and Abdul Samad, Salina and Abid, Ali O. and Hussain, Noor Aini (2008) Robust digit recognition with dynamic time warping and recursive least squares. In: International Symposium on Information Technology, 26-29 Aug. 2008, Kuala Lumpur, Malaysia. (pp. 1-8). 10.1109/ITSIM.2008.4631680
spellingShingle Al-Haddad, Syed Abdul Rahman
Ishak, Khairul Anuar
Abdul Samad, Salina
Abid, Ali O.
Hussain, Noor Aini
Robust digit recognition with dynamic time warping and recursive least squares
title Robust digit recognition with dynamic time warping and recursive least squares
title_full Robust digit recognition with dynamic time warping and recursive least squares
title_fullStr Robust digit recognition with dynamic time warping and recursive least squares
title_full_unstemmed Robust digit recognition with dynamic time warping and recursive least squares
title_short Robust digit recognition with dynamic time warping and recursive least squares
title_sort robust digit recognition with dynamic time warping and recursive least squares
url http://psasir.upm.edu.my/id/eprint/16587/
http://psasir.upm.edu.my/id/eprint/16587/