Code-Mixed Street Address Recognition and Accent Adaptation for Voice-Activated Navigation Services

<p dir="ltr">This study presents the development of a real-time application-specific Automatic Speech Recognition (ASR) system for voice-activated navigation services. The system is designed to recognize Urdu-English code-mixed street addresses, which is challenging due to their comp...

وصف كامل

محفوظ في:
التفاصيل البيبلوغرافية
المؤلف الرئيسي: Syed Meesam Raza Naqvi (22225462) (author)
مؤلفون آخرون: Muhammad Ali Tahir (6640979) (author), Kamran Javed (21726248) (author), Hassan Aqeel Khan (22225465) (author), Ali Raza (3558965) (author), Zubair Saeed (19325647) (author)
منشور في: 2024
الموضوعات:
الوسوم: إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
_version_ 1864513539945791488
author Syed Meesam Raza Naqvi (22225462)
author2 Muhammad Ali Tahir (6640979)
Kamran Javed (21726248)
Hassan Aqeel Khan (22225465)
Ali Raza (3558965)
Zubair Saeed (19325647)
author2_role author
author
author
author
author
author_facet Syed Meesam Raza Naqvi (22225462)
Muhammad Ali Tahir (6640979)
Kamran Javed (21726248)
Hassan Aqeel Khan (22225465)
Ali Raza (3558965)
Zubair Saeed (19325647)
author_role author
dc.creator.none.fl_str_mv Syed Meesam Raza Naqvi (22225462)
Muhammad Ali Tahir (6640979)
Kamran Javed (21726248)
Hassan Aqeel Khan (22225465)
Ali Raza (3558965)
Zubair Saeed (19325647)
dc.date.none.fl_str_mv 2024-11-22T15:00:00Z
dc.identifier.none.fl_str_mv 10.1109/access.2024.3496617
dc.relation.none.fl_str_mv https://figshare.com/articles/journal_contribution/Code-Mixed_Street_Address_Recognition_and_Accent_Adaptation_for_Voice-Activated_Navigation_Services/30095377
dc.rights.none.fl_str_mv CC BY 4.0
info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Engineering
Electronics, sensors and digital hardware
Information and computing sciences
Machine learning
Urdu-English code-mixing
roman Urdu addresses
accent adaptation
deep neural network
Gaussian mixture models
hidden Markov models
voice-activated navigation
Speech recognition
Acoustics
Vocabulary
Speech coding
Real-time systems
Navigation
Long short term memory
Error analysis
Switches
Natural language processing
dc.title.none.fl_str_mv Code-Mixed Street Address Recognition and Accent Adaptation for Voice-Activated Navigation Services
dc.type.none.fl_str_mv Text
Journal contribution
info:eu-repo/semantics/publishedVersion
text
contribution to journal
description <p dir="ltr">This study presents the development of a real-time application-specific Automatic Speech Recognition (ASR) system for voice-activated navigation services. The system is designed to recognize Urdu-English code-mixed street addresses, which is challenging due to their complex nature and structure, especially in under-resourced languages such as Urdu. Two separate corpora are collected for ASR system development: Unicode Urdu consisting of general Urdu recordings of around 61.82 hours by 144 speakers and Roman Urdu-English code-mixed Addresses of around 16.89 hours by 20 speakers. The Unicode Urdu data is developed to provide acoustic models with general language understanding and code-mixed street addresses to provide code-mixing or switching coverage. The hybrid ASR system employed in this study plays a crucial role in addressing the multifaceted challenges of low-resource settings (only 16.89 hours of task-specific data), especially in the context of Urdu-English code-switching. The study compares various acoustic models, with mixed Time Delay Neural Network and Long Short-Term Memory (TDNN-LSTM) performing best with a Word Error Rate (WER), Character Error Rate (CER), and Sentence Error Rate (SER) of 4.02%, 0.8%, and 15.14% respectively, on random street addresses. In addition to testing street addresses, we performed accent-based and manual decoding testing on the developed ASR system. Results indicate the need to develop and deploy custom ASR systems for better accent adaptation and application-specific coverage.</p><h2>Other Information</h2><p dir="ltr">Published in: IEEE Access<br>License: <a href="https://creativecommons.org/licenses/by/4.0/deed.en" target="_blank">https://creativecommons.org/licenses/by/4.0/</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.1109/access.2024.3496617" target="_blank">https://dx.doi.org/10.1109/access.2024.3496617</a></p>
eu_rights_str_mv openAccess
id Manara2_641be73879a2241d48947d16557b7dcc
identifier_str_mv 10.1109/access.2024.3496617
network_acronym_str Manara2
network_name_str Manara2
oai_identifier_str oai:figshare.com:article/30095377
publishDate 2024
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
rights_invalid_str_mv CC BY 4.0
spelling Code-Mixed Street Address Recognition and Accent Adaptation for Voice-Activated Navigation ServicesSyed Meesam Raza Naqvi (22225462)Muhammad Ali Tahir (6640979)Kamran Javed (21726248)Hassan Aqeel Khan (22225465)Ali Raza (3558965)Zubair Saeed (19325647)EngineeringElectronics, sensors and digital hardwareInformation and computing sciencesMachine learningUrdu-English code-mixingroman Urdu addressesaccent adaptationdeep neural networkGaussian mixture modelshidden Markov modelsvoice-activated navigationSpeech recognitionAcousticsVocabularySpeech codingReal-time systemsNavigationLong short term memoryError analysisSwitchesNatural language processing<p dir="ltr">This study presents the development of a real-time application-specific Automatic Speech Recognition (ASR) system for voice-activated navigation services. The system is designed to recognize Urdu-English code-mixed street addresses, which is challenging due to their complex nature and structure, especially in under-resourced languages such as Urdu. Two separate corpora are collected for ASR system development: Unicode Urdu consisting of general Urdu recordings of around 61.82 hours by 144 speakers and Roman Urdu-English code-mixed Addresses of around 16.89 hours by 20 speakers. The Unicode Urdu data is developed to provide acoustic models with general language understanding and code-mixed street addresses to provide code-mixing or switching coverage. The hybrid ASR system employed in this study plays a crucial role in addressing the multifaceted challenges of low-resource settings (only 16.89 hours of task-specific data), especially in the context of Urdu-English code-switching. The study compares various acoustic models, with mixed Time Delay Neural Network and Long Short-Term Memory (TDNN-LSTM) performing best with a Word Error Rate (WER), Character Error Rate (CER), and Sentence Error Rate (SER) of 4.02%, 0.8%, and 15.14% respectively, on random street addresses. In addition to testing street addresses, we performed accent-based and manual decoding testing on the developed ASR system. Results indicate the need to develop and deploy custom ASR systems for better accent adaptation and application-specific coverage.</p><h2>Other Information</h2><p dir="ltr">Published in: IEEE Access<br>License: <a href="https://creativecommons.org/licenses/by/4.0/deed.en" target="_blank">https://creativecommons.org/licenses/by/4.0/</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.1109/access.2024.3496617" target="_blank">https://dx.doi.org/10.1109/access.2024.3496617</a></p>2024-11-22T15:00:00ZTextJournal contributioninfo:eu-repo/semantics/publishedVersiontextcontribution to journal10.1109/access.2024.3496617https://figshare.com/articles/journal_contribution/Code-Mixed_Street_Address_Recognition_and_Accent_Adaptation_for_Voice-Activated_Navigation_Services/30095377CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/300953772024-11-22T15:00:00Z
spellingShingle Code-Mixed Street Address Recognition and Accent Adaptation for Voice-Activated Navigation Services
Syed Meesam Raza Naqvi (22225462)
Engineering
Electronics, sensors and digital hardware
Information and computing sciences
Machine learning
Urdu-English code-mixing
roman Urdu addresses
accent adaptation
deep neural network
Gaussian mixture models
hidden Markov models
voice-activated navigation
Speech recognition
Acoustics
Vocabulary
Speech coding
Real-time systems
Navigation
Long short term memory
Error analysis
Switches
Natural language processing
status_str publishedVersion
title Code-Mixed Street Address Recognition and Accent Adaptation for Voice-Activated Navigation Services
title_full Code-Mixed Street Address Recognition and Accent Adaptation for Voice-Activated Navigation Services
title_fullStr Code-Mixed Street Address Recognition and Accent Adaptation for Voice-Activated Navigation Services
title_full_unstemmed Code-Mixed Street Address Recognition and Accent Adaptation for Voice-Activated Navigation Services
title_short Code-Mixed Street Address Recognition and Accent Adaptation for Voice-Activated Navigation Services
title_sort Code-Mixed Street Address Recognition and Accent Adaptation for Voice-Activated Navigation Services
topic Engineering
Electronics, sensors and digital hardware
Information and computing sciences
Machine learning
Urdu-English code-mixing
roman Urdu addresses
accent adaptation
deep neural network
Gaussian mixture models
hidden Markov models
voice-activated navigation
Speech recognition
Acoustics
Vocabulary
Speech coding
Real-time systems
Navigation
Long short term memory
Error analysis
Switches
Natural language processing