Automatic Speech Recognition Method Based on Deep Learning Approaches for Uzbek Language
Communication has been an important aspect of human life, civilization, and globalization for thousands of years. Biometric analysis, education, security, healthcare, and smart cities are only a few examples of speech recognition applications. Most studies have mainly concentrated on English, Spanis...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-05-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/22/10/3683 |
| _version_ | 1827666686134714368 |
|---|---|
| author | Abdinabi Mukhamadiyev Ilyos Khujayarov Oybek Djuraev Jinsoo Cho |
| author_facet | Abdinabi Mukhamadiyev Ilyos Khujayarov Oybek Djuraev Jinsoo Cho |
| author_sort | Abdinabi Mukhamadiyev |
| collection | DOAJ |
| description | Communication has been an important aspect of human life, civilization, and globalization for thousands of years. Biometric analysis, education, security, healthcare, and smart cities are only a few examples of speech recognition applications. Most studies have mainly concentrated on English, Spanish, Japanese, or Chinese, disregarding other low-resource languages, such as Uzbek, leaving their analysis open. In this paper, we propose an End-To-End Deep Neural Network-Hidden Markov Model speech recognition model and a hybrid Connectionist Temporal Classification (CTC)-attention network for the Uzbek language and its dialects. The proposed approach reduces training time and improves speech recognition accuracy by effectively using CTC objective function in attention model training. We evaluated the linguistic and lay-native speaker performances on the Uzbek language dataset, which was collected as a part of this study. Experimental results show that the proposed model achieved a word error rate of 14.3% using 207 h of recordings as an Uzbek language training dataset. |
| first_indexed | 2024-03-10T01:55:11Z |
| format | Article |
| id | doaj.art-a252ecac5001421f925c3344425a9384 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T01:55:11Z |
| publishDate | 2022-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-a252ecac5001421f925c3344425a93842023-11-23T12:59:23ZengMDPI AGSensors1424-82202022-05-012210368310.3390/s22103683Automatic Speech Recognition Method Based on Deep Learning Approaches for Uzbek LanguageAbdinabi Mukhamadiyev0Ilyos Khujayarov1Oybek Djuraev2Jinsoo Cho3Department of Computer Engineering, Gachon University, Sujeong-gu, Seongnam-si 13120, KoreaDepartment of Information Technologies, Samarkand Branch of Tashkent University of Information Technologies Named after Muhammad al-Khwarizmi, Tashkent 140100, UzbekistanDepartment of Hardware and Software of Control Systems in Telecommunication, Tashkent University of Information Technologies Named after Muhammad al-Khwarizmi, Tashkent 100084, UzbekistanDepartment of Computer Engineering, Gachon University, Sujeong-gu, Seongnam-si 13120, KoreaCommunication has been an important aspect of human life, civilization, and globalization for thousands of years. Biometric analysis, education, security, healthcare, and smart cities are only a few examples of speech recognition applications. Most studies have mainly concentrated on English, Spanish, Japanese, or Chinese, disregarding other low-resource languages, such as Uzbek, leaving their analysis open. In this paper, we propose an End-To-End Deep Neural Network-Hidden Markov Model speech recognition model and a hybrid Connectionist Temporal Classification (CTC)-attention network for the Uzbek language and its dialects. The proposed approach reduces training time and improves speech recognition accuracy by effectively using CTC objective function in attention model training. We evaluated the linguistic and lay-native speaker performances on the Uzbek language dataset, which was collected as a part of this study. Experimental results show that the proposed model achieved a word error rate of 14.3% using 207 h of recordings as an Uzbek language training dataset.https://www.mdpi.com/1424-8220/22/10/3683convolutional neural networkend-to-end speech recognitiontransformersCTC-attentionUzbek languagedeep learning |
| spellingShingle | Abdinabi Mukhamadiyev Ilyos Khujayarov Oybek Djuraev Jinsoo Cho Automatic Speech Recognition Method Based on Deep Learning Approaches for Uzbek Language Sensors convolutional neural network end-to-end speech recognition transformers CTC-attention Uzbek language deep learning |
| title | Automatic Speech Recognition Method Based on Deep Learning Approaches for Uzbek Language |
| title_full | Automatic Speech Recognition Method Based on Deep Learning Approaches for Uzbek Language |
| title_fullStr | Automatic Speech Recognition Method Based on Deep Learning Approaches for Uzbek Language |
| title_full_unstemmed | Automatic Speech Recognition Method Based on Deep Learning Approaches for Uzbek Language |
| title_short | Automatic Speech Recognition Method Based on Deep Learning Approaches for Uzbek Language |
| title_sort | automatic speech recognition method based on deep learning approaches for uzbek language |
| topic | convolutional neural network end-to-end speech recognition transformers CTC-attention Uzbek language deep learning |
| url | https://www.mdpi.com/1424-8220/22/10/3683 |
| work_keys_str_mv | AT abdinabimukhamadiyev automaticspeechrecognitionmethodbasedondeeplearningapproachesforuzbeklanguage AT ilyoskhujayarov automaticspeechrecognitionmethodbasedondeeplearningapproachesforuzbeklanguage AT oybekdjuraev automaticspeechrecognitionmethodbasedondeeplearningapproachesforuzbeklanguage AT jinsoocho automaticspeechrecognitionmethodbasedondeeplearningapproachesforuzbeklanguage |