Improvement of speech recognition results by a combination of systems

Rama Hasan, Hussein Hussein, Pavlos Lazaridis, Sinan Khwandah, Marc Ritter, Maximilian Eibl

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

The aim of this study is to suggest an algorithm that combines two speech recognition systems. These systems differ in the methods used in the feature extraction stage, but they have the same classifier Hidden Markov Model (HMM). The first system uses Mel-Frequency Cepstrum Coefficients (MFCC), the second one uses Linear Prediction Cepstrum Coefficients (LPCC), and the third system uses Perceptual Linear Predictive (PLP) features. The combination algorithm is applied separately on each couple of systems. The study is implemented on a data set that consists of the four voice commands: 'shutdown', 'documents', 'restart', and 'net' pronounced by 33 people. In addition to the improvement of the speech recognition rate for isolated words, the study aimed to determine the most complementary couple of systems through studying two kinds of errors: simultaneous and dependent errors. The system depending on MFCC features provided the highest recognition rate with 85.44%. The results showed noticeable improvement of combined systems in comparison with the individual systems where combining MFCC & PLP provided the highest recognition rate with 93.44%.

LanguageEnglish
Title of host publication2017 23rd IEEE International Conference on Automation and Computing (ICAC)
Subtitle of host publicationAddressing Global Challenges through Automation and Computing
EditorsJie Zhang
PublisherInstitute of Electrical and Electronics Engineers Inc.
Number of pages4
ISBN (Electronic)9780701702601
DOIs
Publication statusPublished - 26 Oct 2017
Event23rd International Conference on Automation and Computing: Addressing Global Challenges through Automation and Computing - University of Huddersfield, Huddersfield, United Kingdom
Duration: 7 Sep 20178 Sep 2017
Conference number: 23
https://www.ieee.org/conferences_events/conferences/conferencedetails/index.html?Conf_ID=41042 (Link to Conference Website)

Conference

Conference23rd International Conference on Automation and Computing
Abbreviated titleICAC 2017
CountryUnited Kingdom
CityHuddersfield
Period7/09/178/09/17
OtherThe scope of the conference covers a broad spectrum of areas with multi-disciplinary interests in the fields of automation, control engineering, computing and information systems, ranging from fundamental research to real-world applications.
Internet address

Fingerprint

Speech Recognition
Speech recognition
Cepstrum
Hidden Markov models
Feature extraction
Classifiers
Coefficient
Linear Prediction
Restart
Markov Model
Feature Extraction
Classifier
Datasets
Dependent

Cite this

Hasan, R., Hussein, H., Lazaridis, P., Khwandah, S., Ritter, M., & Eibl, M. (2017). Improvement of speech recognition results by a combination of systems. In J. Zhang (Ed.), 2017 23rd IEEE International Conference on Automation and Computing (ICAC): Addressing Global Challenges through Automation and Computing [8082082] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.23919/IConAC.2017.8082082
Hasan, Rama ; Hussein, Hussein ; Lazaridis, Pavlos ; Khwandah, Sinan ; Ritter, Marc ; Eibl, Maximilian. / Improvement of speech recognition results by a combination of systems. 2017 23rd IEEE International Conference on Automation and Computing (ICAC): Addressing Global Challenges through Automation and Computing. editor / Jie Zhang. Institute of Electrical and Electronics Engineers Inc., 2017.
@inproceedings{6959845aacbc47b9a3b7985f94b515de,
title = "Improvement of speech recognition results by a combination of systems",
abstract = "The aim of this study is to suggest an algorithm that combines two speech recognition systems. These systems differ in the methods used in the feature extraction stage, but they have the same classifier Hidden Markov Model (HMM). The first system uses Mel-Frequency Cepstrum Coefficients (MFCC), the second one uses Linear Prediction Cepstrum Coefficients (LPCC), and the third system uses Perceptual Linear Predictive (PLP) features. The combination algorithm is applied separately on each couple of systems. The study is implemented on a data set that consists of the four voice commands: 'shutdown', 'documents', 'restart', and 'net' pronounced by 33 people. In addition to the improvement of the speech recognition rate for isolated words, the study aimed to determine the most complementary couple of systems through studying two kinds of errors: simultaneous and dependent errors. The system depending on MFCC features provided the highest recognition rate with 85.44{\%}. The results showed noticeable improvement of combined systems in comparison with the individual systems where combining MFCC & PLP provided the highest recognition rate with 93.44{\%}.",
keywords = "Features extraction, Hidden Markov models, Speech recognition",
author = "Rama Hasan and Hussein Hussein and Pavlos Lazaridis and Sinan Khwandah and Marc Ritter and Maximilian Eibl",
year = "2017",
month = "10",
day = "26",
doi = "10.23919/IConAC.2017.8082082",
language = "English",
editor = "Jie Zhang",
booktitle = "2017 23rd IEEE International Conference on Automation and Computing (ICAC)",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

Hasan, R, Hussein, H, Lazaridis, P, Khwandah, S, Ritter, M & Eibl, M 2017, Improvement of speech recognition results by a combination of systems. in J Zhang (ed.), 2017 23rd IEEE International Conference on Automation and Computing (ICAC): Addressing Global Challenges through Automation and Computing., 8082082, Institute of Electrical and Electronics Engineers Inc., 23rd International Conference on Automation and Computing, Huddersfield, United Kingdom, 7/09/17. https://doi.org/10.23919/IConAC.2017.8082082

Improvement of speech recognition results by a combination of systems. / Hasan, Rama; Hussein, Hussein; Lazaridis, Pavlos; Khwandah, Sinan; Ritter, Marc; Eibl, Maximilian.

2017 23rd IEEE International Conference on Automation and Computing (ICAC): Addressing Global Challenges through Automation and Computing. ed. / Jie Zhang. Institute of Electrical and Electronics Engineers Inc., 2017. 8082082.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Improvement of speech recognition results by a combination of systems

AU - Hasan, Rama

AU - Hussein, Hussein

AU - Lazaridis, Pavlos

AU - Khwandah, Sinan

AU - Ritter, Marc

AU - Eibl, Maximilian

PY - 2017/10/26

Y1 - 2017/10/26

N2 - The aim of this study is to suggest an algorithm that combines two speech recognition systems. These systems differ in the methods used in the feature extraction stage, but they have the same classifier Hidden Markov Model (HMM). The first system uses Mel-Frequency Cepstrum Coefficients (MFCC), the second one uses Linear Prediction Cepstrum Coefficients (LPCC), and the third system uses Perceptual Linear Predictive (PLP) features. The combination algorithm is applied separately on each couple of systems. The study is implemented on a data set that consists of the four voice commands: 'shutdown', 'documents', 'restart', and 'net' pronounced by 33 people. In addition to the improvement of the speech recognition rate for isolated words, the study aimed to determine the most complementary couple of systems through studying two kinds of errors: simultaneous and dependent errors. The system depending on MFCC features provided the highest recognition rate with 85.44%. The results showed noticeable improvement of combined systems in comparison with the individual systems where combining MFCC & PLP provided the highest recognition rate with 93.44%.

AB - The aim of this study is to suggest an algorithm that combines two speech recognition systems. These systems differ in the methods used in the feature extraction stage, but they have the same classifier Hidden Markov Model (HMM). The first system uses Mel-Frequency Cepstrum Coefficients (MFCC), the second one uses Linear Prediction Cepstrum Coefficients (LPCC), and the third system uses Perceptual Linear Predictive (PLP) features. The combination algorithm is applied separately on each couple of systems. The study is implemented on a data set that consists of the four voice commands: 'shutdown', 'documents', 'restart', and 'net' pronounced by 33 people. In addition to the improvement of the speech recognition rate for isolated words, the study aimed to determine the most complementary couple of systems through studying two kinds of errors: simultaneous and dependent errors. The system depending on MFCC features provided the highest recognition rate with 85.44%. The results showed noticeable improvement of combined systems in comparison with the individual systems where combining MFCC & PLP provided the highest recognition rate with 93.44%.

KW - Features extraction

KW - Hidden Markov models

KW - Speech recognition

UR - http://www.scopus.com/inward/record.url?scp=85039989988&partnerID=8YFLogxK

U2 - 10.23919/IConAC.2017.8082082

DO - 10.23919/IConAC.2017.8082082

M3 - Conference contribution

BT - 2017 23rd IEEE International Conference on Automation and Computing (ICAC)

A2 - Zhang, Jie

PB - Institute of Electrical and Electronics Engineers Inc.

ER -

Hasan R, Hussein H, Lazaridis P, Khwandah S, Ritter M, Eibl M. Improvement of speech recognition results by a combination of systems. In Zhang J, editor, 2017 23rd IEEE International Conference on Automation and Computing (ICAC): Addressing Global Challenges through Automation and Computing. Institute of Electrical and Electronics Engineers Inc. 2017. 8082082 https://doi.org/10.23919/IConAC.2017.8082082