The Study of Malay’s Prosodic Features Impact on Classical Arabic Accents Recognition
Modeling individual’s variation in speech pattern can be challenging in Automatic Speech Recognition (ASR). In Classical Arabic (CA) language, 20 Quranic accents are permitted for Quranic recitation. An ASR system for CA with accent detection requires a modeling method that can capture sp...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2023-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/10196429/ |
_version_ | 1797688449860370432 |
---|---|
author | Noor Jamaliah Ibrahim Mohd Yamani Idna Idris M. Y. Zulkifli Mohd Yusoff Roziana Ramli Raja Jamilah Raja Yusof |
author_facet | Noor Jamaliah Ibrahim Mohd Yamani Idna Idris M. Y. Zulkifli Mohd Yusoff Roziana Ramli Raja Jamilah Raja Yusof |
author_sort | Noor Jamaliah Ibrahim |
collection | DOAJ |
description | Modeling individual’s variation in speech pattern can be challenging in Automatic Speech Recognition (ASR). In Classical Arabic (CA) language, 20 Quranic accents are permitted for Quranic recitation. An ASR system for CA with accent detection requires a modeling method that can capture speech pattern changes. Here, we study the accentual influences on Malay speakers’ pronunciation and its prosodic impacts towards ASR system for CA language with seven Quranic accents identification. The proposed ASR system was developed over three stages. First, a dataset of Surah Al-Fatihah recitation was recorded from 14 Malay speakers in seven Quranic accents, forming a total of 5,684 words. Second, various spectral and prosodic features are extracted from the dataset for further classification process. The final stage includes training and testing the classification model. The existing ASR systems are often enabled by Gaussian Mixture Models (GMM) because of its capability to represent a wide range of sample distributions. However, GMM is susceptible to overfitting when the model complexity is high, due to the presence of singularities. To support identification of seven Quranic accents, Universal Background Model (UBM) is adapted to GMM using Maximum A Posteriori (MAP) estimation method. The UBM models were trained over each of Quranic accents, and combined to establish final UBM with 512 mixture components. The proposed ASR system utilizing the GMM-UBM outperformed k-NN, GMM, and GMM-iVector in identifying Al-Fatihah recitation to the corresponding Quranic accents. The GMM-UBM yields a testing accuracy of 86.148%, which is an increment of 4.435% from utilizing GMM alone. |
first_indexed | 2024-03-12T01:32:09Z |
format | Article |
id | doaj.art-5240ee5be85b44dbbb6bb0ebe1dd0820 |
institution | Directory Open Access Journal |
issn | 2169-3536 |
language | English |
last_indexed | 2024-03-12T01:32:09Z |
publishDate | 2023-01-01 |
publisher | IEEE |
record_format | Article |
series | IEEE Access |
spelling | doaj.art-5240ee5be85b44dbbb6bb0ebe1dd08202023-09-11T23:02:10ZengIEEEIEEE Access2169-35362023-01-0111945899461210.1109/ACCESS.2023.329981410196429The Study of Malay’s Prosodic Features Impact on Classical Arabic Accents RecognitionNoor Jamaliah Ibrahim0https://orcid.org/0009-0008-9600-3869Mohd Yamani Idna Idris1https://orcid.org/0000-0003-4894-0838M. Y. Zulkifli Mohd Yusoff2Roziana Ramli3https://orcid.org/0000-0002-3763-3149Raja Jamilah Raja Yusof4https://orcid.org/0000-0001-9894-1893Department of Computer System and Technology, Faculty of Computer Science and Information Technology, Universiti Malaya, Kuala Lumpur, MalaysiaDepartment of Computer System and Technology, Faculty of Computer Science and Information Technology, Universiti Malaya, Kuala Lumpur, MalaysiaDepartment of Al-Quran and Al-Hadith, Academy of Islamic Studies, Universiti Malaya, Kuala Lumpur, MalaysiaDepartment of Computer System and Technology, Faculty of Computer Science and Information Technology, Universiti Malaya, Kuala Lumpur, MalaysiaDepartment of Software Engineering, Faculty of Computer Science and Information Technology, Universiti Malaya, Kuala Lumpur, MalaysiaModeling individual’s variation in speech pattern can be challenging in Automatic Speech Recognition (ASR). In Classical Arabic (CA) language, 20 Quranic accents are permitted for Quranic recitation. An ASR system for CA with accent detection requires a modeling method that can capture speech pattern changes. Here, we study the accentual influences on Malay speakers’ pronunciation and its prosodic impacts towards ASR system for CA language with seven Quranic accents identification. The proposed ASR system was developed over three stages. First, a dataset of Surah Al-Fatihah recitation was recorded from 14 Malay speakers in seven Quranic accents, forming a total of 5,684 words. Second, various spectral and prosodic features are extracted from the dataset for further classification process. The final stage includes training and testing the classification model. The existing ASR systems are often enabled by Gaussian Mixture Models (GMM) because of its capability to represent a wide range of sample distributions. However, GMM is susceptible to overfitting when the model complexity is high, due to the presence of singularities. To support identification of seven Quranic accents, Universal Background Model (UBM) is adapted to GMM using Maximum A Posteriori (MAP) estimation method. The UBM models were trained over each of Quranic accents, and combined to establish final UBM with 512 mixture components. The proposed ASR system utilizing the GMM-UBM outperformed k-NN, GMM, and GMM-iVector in identifying Al-Fatihah recitation to the corresponding Quranic accents. The GMM-UBM yields a testing accuracy of 86.148%, which is an increment of 4.435% from utilizing GMM alone.https://ieeexplore.ieee.org/document/10196429/Automatic speech recognition (ASR)Gaussian mixture model-universal background model (GMM-UBM)Malay speakersQuranic accents |
spellingShingle | Noor Jamaliah Ibrahim Mohd Yamani Idna Idris M. Y. Zulkifli Mohd Yusoff Roziana Ramli Raja Jamilah Raja Yusof The Study of Malay’s Prosodic Features Impact on Classical Arabic Accents Recognition IEEE Access Automatic speech recognition (ASR) Gaussian mixture model-universal background model (GMM-UBM) Malay speakers Quranic accents |
title | The Study of Malay’s Prosodic Features Impact on Classical Arabic Accents Recognition |
title_full | The Study of Malay’s Prosodic Features Impact on Classical Arabic Accents Recognition |
title_fullStr | The Study of Malay’s Prosodic Features Impact on Classical Arabic Accents Recognition |
title_full_unstemmed | The Study of Malay’s Prosodic Features Impact on Classical Arabic Accents Recognition |
title_short | The Study of Malay’s Prosodic Features Impact on Classical Arabic Accents Recognition |
title_sort | study of malay x2019 s prosodic features impact on classical arabic accents recognition |
topic | Automatic speech recognition (ASR) Gaussian mixture model-universal background model (GMM-UBM) Malay speakers Quranic accents |
url | https://ieeexplore.ieee.org/document/10196429/ |
work_keys_str_mv | AT noorjamaliahibrahim thestudyofmalayx2019sprosodicfeaturesimpactonclassicalarabicaccentsrecognition AT mohdyamaniidnaidris thestudyofmalayx2019sprosodicfeaturesimpactonclassicalarabicaccentsrecognition AT myzulkiflimohdyusoff thestudyofmalayx2019sprosodicfeaturesimpactonclassicalarabicaccentsrecognition AT rozianaramli thestudyofmalayx2019sprosodicfeaturesimpactonclassicalarabicaccentsrecognition AT rajajamilahrajayusof thestudyofmalayx2019sprosodicfeaturesimpactonclassicalarabicaccentsrecognition AT noorjamaliahibrahim studyofmalayx2019sprosodicfeaturesimpactonclassicalarabicaccentsrecognition AT mohdyamaniidnaidris studyofmalayx2019sprosodicfeaturesimpactonclassicalarabicaccentsrecognition AT myzulkiflimohdyusoff studyofmalayx2019sprosodicfeaturesimpactonclassicalarabicaccentsrecognition AT rozianaramli studyofmalayx2019sprosodicfeaturesimpactonclassicalarabicaccentsrecognition AT rajajamilahrajayusof studyofmalayx2019sprosodicfeaturesimpactonclassicalarabicaccentsrecognition |