Melola Discovery

← Melola Discovery

Jurnal SINTA 1

Fine-Tuning Whisper Model for Mandar Speech Recognition: Approach and Performance Evaluation

Journal of Applied Engineering and Technological Science (JAETS) · Vol. 7 No. 1 (2025) · hlm. 261–273

DOI
10.37385/jaets.v7i1.7170
ISSN
2715-6079, 2715-6087
Terbit
2025-12-29
Penerbit
Yayasan Riset dan Pengembangan Intelektual (YRPI)
Bahasa
en

Abstrak

This research focuses on the development of speech recognition technology for the Mandar language, a regional language in Indonesia with limited digital resources. The main challenge lies in the lack of local datasets and the minimal representation of the Mandar language in existing multilingual speech recognition models. This study aims to enhance the performance of Automatic Speech Recognition (ASR) systems by fine-tuning the Whisper model using a Mandar-specific dataset. The dataset consists of 1,000 audio recordings with various dialects and recording qualities, which underwent preprocessing steps such as segmentation, normalization, and data augmentation. Fine-tuning was conducted using supervised learning methods with hyperparameter optimization, resulting in a reduction of Word Error Rate (WER) from 73.7% in the pretrained model to 37.4% after fine-tuning, and an increase in accuracy from 26.3% to 62.6%. The optimized model was also compared with other ASR models, such as DeepSpeech and wav2vec 2.0, demonstrating superior performance in terms of accuracy and time efficiency. Further analysis revealed that recording quality and dialect variations significantly impacted model performance, with high-quality recordings and standard dialects yielding the best results. The model was implemented as a web application prototype, enabling efficient and near real-time transcription of Mandar speech. This research not only contributes to the development of ASR technology for low-resource languages but also opens new opportunities for preserving and utilizing the Mandar language through digital technology. For future improvements, larger datasets, more advanced augmentation techniques, and the exploration of additional language model integration are recommended.

Kata kunci

Jaringan kolaborasi para penulis

20 penulis · 54 tautan kepenulisan bersama · 28 lintas organisasi

Arahkan kursor atau ketuk sebuah nama untuk menyorot rekan penulisnya. Klik dua kali untuk membuka halaman penulis.

  • Universitas Pancasakti Makassar
  • State University of Makassar
  • Lambung Mangkurat University
  • Hasanuddin University
  • Institut Teknologi Bacharuddin Jusuf Habibie
  • Cenderawasih University
  • Sekolah Tinggi Ilmu Ekonomi Makassar Bongaya
  • Afiliasi tidak diketahui
  • Organisasi berbeda
  • Organisasi sama
  • Afiliasi tidak diketahui
  • Garis lebih tebal = lebih banyak artikel bersama · lingkaran lebih besar = lebih banyak artikel
Tampilkan sebagai tabel
Rekan penulisOrganisasiArtikel bersama
Supriyadi La WungoAfiliasi tidak diketahui2
Nurul Fathanah MustaminLambung Mangkurat University2
Jeffry JeffryInstitut Teknologi Bacharuddin Jusuf Habibie2
Mutia MaulidaLambung Mangkurat University2
Jafar JafarUniversitas Pancasakti Makassar2
Nurafni ShahnybUniversitas Pancasakti Makassar2
Ampauleng AmpaulengSekolah Tinggi Ilmu Ekonomi Makassar Bongaya2
Firmansyah FirmansyahHasanuddin University1
Pertiwi IshakUniversitas Pancasakti Makassar1
Armin LawiHasanuddin University1
Misbahuddin BuangAfiliasi tidak diketahui1
Nasruddin NasruddinCenderawasih University1
Syahrul UsmanUniversitas Pancasakti Makassar1
Rahmat Fuadi SyamUniversitas Pancasakti Makassar1
Muhammad Nur ArafahAfiliasi tidak diketahui1

Afiliasi berasal dari OpenAlex; metadata OAI-PMH jurnal tidak memuatnya. Penulis tanpa ID OpenAlex dicocokkan berdasarkan nama, sehingga dua orang bisa tergabung atau satu orang bisa terpisah.

Sumber metadata: OAI-PMH (metadata jurnal) https://journal.yrpipku.com/index.php/jaets/oai, pertama di-harvest 9 Nov 2026, terakhir terlihat 9 Nov 2026.
Metadata asli (oai_dc)
<oai_dc:dc xmlns="http://www.openarchives.org/OAI/2.0/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:oai_dc="http://www.openarchives.org/OAI/2.0/oai_dc/" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/oai_dc/  http://www.openarchives.org/OAI/2.0/oai_dc.xsd">
	<dc:title xml:lang="en">Fine-Tuning Whisper Model for Mandar Speech Recognition: Approach and Performance Evaluation</dc:title>
	<dc:creator xml:lang="en">Jafar, Jafar</dc:creator>
	<dc:creator xml:lang="en">Tb, Mar Athul Wazithah</dc:creator>
	<dc:creator xml:lang="en">Aziz, Firman</dc:creator>
	<dc:creator xml:lang="en">Iriany, Rosary</dc:creator>
	<dc:creator xml:lang="en">Nasir, Norma</dc:creator>
	<dc:subject xml:lang="en">Whisper</dc:subject>
	<dc:subject xml:lang="en">Fine-Tuning</dc:subject>
	<dc:subject xml:lang="en">Mandar Language</dc:subject>
	<dc:subject xml:lang="en">Word Error Rate (WER)</dc:subject>
	<dc:subject xml:lang="en">Automatic Speech Recognition</dc:subject>
	<dc:description xml:lang="en">This research focuses on the development of speech recognition technology for the Mandar language, a regional language in Indonesia with limited digital resources. The main challenge lies in the lack of local datasets and the minimal representation of the Mandar language in existing multilingual speech recognition models. This study aims to enhance the performance of Automatic Speech Recognition (ASR) systems by fine-tuning the Whisper model using a Mandar-specific dataset. The dataset consists of 1,000 audio recordings with various dialects and recording qualities, which underwent preprocessing steps such as segmentation, normalization, and data augmentation. Fine-tuning was conducted using supervised learning methods with hyperparameter optimization, resulting in a reduction of Word Error Rate (WER) from 73.7% in the pretrained model to 37.4% after fine-tuning, and an increase in accuracy from 26.3% to 62.6%. The optimized model was also compared with other ASR models, such as DeepSpeech and wav2vec 2.0, demonstrating superior performance in terms of accuracy and time efficiency. Further analysis revealed that recording quality and dialect variations significantly impacted model performance, with high-quality recordings and standard dialects yielding the best results. The model was implemented as a web application prototype, enabling efficient and near real-time transcription of Mandar speech. This research not only contributes to the development of ASR technology for low-resource languages but also opens new opportunities for preserving and utilizing the Mandar language through digital technology. For future improvements, larger datasets, more advanced augmentation techniques, and the exploration of additional language model integration are recommended.</dc:description>
	<dc:publisher xml:lang="en">Yayasan Riset dan Pengembangan Intelektual (YRPI)</dc:publisher>
	<dc:date>2025-12-29</dc:date>
	<dc:type>info:eu-repo/semantics/article</dc:type>
	<dc:type>info:eu-repo/semantics/publishedVersion</dc:type>
	<dc:format>application/pdf</dc:format>
	<dc:identifier>https://journal.yrpipku.com/index.php/jaets/article/view/7170</dc:identifier>
	<dc:identifier>10.37385/jaets.v7i1.7170</dc:identifier>
	<dc:source xml:lang="en">Journal of Applied Engineering and Technological Science (JAETS); Vol. 7 No. 1 (2025): Journal of Applied Engineering and Technological Science (JAETS); 261-273</dc:source>
	<dc:source>2715-6079</dc:source>
	<dc:source>2715-6087</dc:source>
	<dc:source>10.37385/jaets.v7i1</dc:source>
	<dc:language>en</dc:language>
	<dc:relation>https://journal.yrpipku.com/index.php/jaets/article/view/7170/5405</dc:relation>
	<dc:rights xml:lang="en">Copyright (c) 2025 Jafar Jafar, Mar Athul Wazithah Tb, Firman Aziz, Rosary Iriany, Norma Nasir</dc:rights>
</oai_dc:dc>