بکارگیری الگوریتم‌های یادگیری واژه‌نامه در بازنمایی تُنُک دادگان گفتاری

نوع مقاله : مقاله پژوهشی

نویسندگان

1 دانشکده فنی و مهندسی، گروه مهندسی برق الکترونیک، دانشگاه لرستان، خرم آباد، ایران

2 دانشکده مهندسی و فناوری، دانشگاه مازندران، بابلسر، ایران

چکیده

بازنمایی تُنُک به عنوان یکی از روش‌های پر کاربرد در پردازش سیگنال، در زمینه‌های مختلفی مانند فشرده‌سازی داده، حذف نویز از سیگنال‌های گفتاری و تصویری، تشخیص الگو و سایر مسائل مرتبط با پردازش سیگنال مورد توجه قرار گرفته است. در چنین بازنمایی‌هایی، سیگنال‌ها با استفاده از تعداد کمی از اتم‌های واژه‌نامه به‌صورت خطی ترکیب می‌شوند که منجر به کاهش ابعاد داده و بهبود کارایی در پردازش سیگنال می‌شود. به منظور بازنمایی دقیق‌تر داده‌های گفتاری، نیاز به واژه‌نامه مناسبی است که بتواند ویژگی‌های سیگنال گفتار را به خوبی نمایش دهد. در این مقاله، واژه‌نامه‌هایی با استفاده از الگوریتم‌های یادگیری واژه‌نامه و بازنمایی تُنُک MOD، K-SVD ،RAMC و UD4-MOD و بازنمایی تُنُک OMP در حوزه‌های زمان، نمایش زمان-فرکانس و تبدیل موجک آموزش داده می‌شوند. ارزیابی کارایی واژه‌نامه‌های به‌دست‌آمده با استفاده از معیارهای مختلف زمانی و فرکانسی مانند RE، MSE، fwSegSNR، SegSNR، PESQ و STOI انجام شده است. نتایج حاصل، نشان می‌دهد که بکارگیری الگوریتم یادگیری واژه‌نامه K-SVD در ترکیب با الگوریتم بازنمایی تُنُک OMP در حوزه STFT نتایج مطلوبی را به منظور بازسازی سیگنال گفتاری به دست می‌دهد.

کلیدواژه‌ها

موضوعات

عنوان مقاله [English]

Applying Dictionary Learning Algorithms In Sparse Representation of Speech Signals

نویسندگان [English]

  • Naser Sharafi 1
  • Salman Kkarimi 1
  • Samira Mavaddati 2
1 Faculty of Engineering, Lorestan University, Khorramabad, Iran
2 Faculty of Engineering and Technology, University of Mazandaran, Babolsar, Iran
چکیده [English]

As a widely used technique in signal processing, Sparse representation has gained significant attention in various fields, including data compression, noise reduction in speech and image signals, pattern recognition, and other signal processing-related issues. In such representations, signals are linearly combined using a small number of dictionary atoms, leading to data dimensionality reduction and improved signal processing efficiency. To accurately represent speech data, an appropriate dictionary is required to effectively represent speech signals' characteristics. In this paper, dictionaries are trained using dictionary learning algorithms and sparse representations such as MOD, K-SVD, RAMC, UD4-MOD, and OMP, in the time, time-frequency, and wavelet transform domains. The performance of the obtained dictionaries is evaluated using various time-frequency metrics such as RE, MSE, fwSegSNR, SegSNR, PESQ, and STOI. The results demonstrate that employing the K-SVD dictionary learning algorithm in conjunction with the OMP sparse representation algorithm in the STFT domain achieves promising results for speech signal reconstruction.

کلیدواژه‌ها [English]

  • Sparse representation
  • Dictionary learning
  • Speech processing
  • K-SVD
  • OMP
  • STFT

مراجع

[1] Tsaig, Yaakov, and David L. Donoho. "Extensions of Compressed Sensing." Signal Processing 86, no. 3 (2006): 549–571.
[2] Zhao, Yongqiang, and Jingxiang Yang. "Hyperspectral Image Denoising via Sparse Representation and Low-Rank Constraint." IEEE Transactions on Geoscience and Remote Sensing 53, no. 1 (2015): 296–308.
[3] Yang, Meng, Dengxin Dai, Linlin Shen, and Luc Van Gool. "Latent Dictionary Learning for Sparse Representation Based Classification." In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 4138–4145. 2014.
[4] Liu, Yu, Xun Chen, Aiping Liu, Rabab K. Ward, and Z. Jane Wang. "Recent Advances in Sparse Representation Based Medical Image Fusion." IEEE Instrumentation & Measurement Magazine 24, no. 2 (2021): 45–53.
[5] Sharma, Pulkit, Vinayak Abrol, and Anil Kumar Sao. "Deep-Sparse-Representation-Based Features for Speech Recognition." IEEE/ACM Transactions on Audio, Speech, and Language Processing 25, no. 11 (2017): 2162–2175.
[6] Peng, Peng, Yongfeng Ju, Yipu Zhang, Kaiming Wang, Suying Jiang, and Yuping Wang. "Sparse Representation and Dictionary Learning Model Incorporating Group Sparsity and Incoherence to Extract Abnormal Brain Regions Associated with Schizophrenia." IEEE Access 8 (2020): 104396–104406.
[7] Tourneret, Jean-Yves, Adrian Basarab, Nora Leïla Ouzir, and Qi Wei. "Sparse Representations and Dictionary Learning: From Image Fusion to Motion Estimation." In 2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS, 25–28. 2021.
[8] Mavaddaty, Samira, Seyed Mohammad Ahadi, and Sanaz Seyedin. "Speech Enhancement Using Sparse Dictionary Learning in Wavelet Packet Transform Domain." Computer Speech & Language 44 (2017): 22–47.
[9] Eshaghi, Mohadese, Farbod Razzazi, and Alireza Behrad. "A Voice Activity Detection Algorithm in Spectro-Temporal Domain Using Sparse Representation." International Journal of Machine Learning and Cybernetics 10, no. 7 (2019): 1791–1803.
[10] Sugiura, Yosuke, and Tetsuya Shimamura. "Speech Enhancement Based on Sparse Representation in Logarithmic Frequency Scale." In 2018 International Symposium on Intelligent Signal Processing and Communication Systems (ISPACS), 252–257. 2018.
[11] Shaheen, Dima, Oumayma Al Dakkak, and Mohiedin Wainakh. "Incoherent Discriminative Dictionary Learning for Speech Enhancement." Journal of Telecommunications and Information Technology (2018): 42–54.
[12] Ji, Yunyun, Wei-Ping Zhu, and Benoit Champagne. "Speech Enhancement Based on Dictionary Learning and Low-Rank Matrix Decomposition." IEEE Access 7 (2019): 4936–4947.
[13] Bai, Huang, Chuanrong Hong, Sheng Li, Yimin D. Zhang, and Xiumei Li. "Unit-Norm Tight Frame-Based Sparse Representation with Application to Speech Inpainting." Digital Signal Processing 123 (2022): 103426.
[14] Wang, Lianzi, Nikos Mastorakis, and Xiaodong Zhuang. "Voiced/Unvoiced Pronunciation Judgement Based on Sparse Representation and Learning Dictionary." In MATEC Web of Conferences 292 (2019): 04012.
[15] Ding, Shaojin, Guanlong Zhao, Christopher Liberatore, and Ricardo Gutierrez-Osuna. "Learning Structured Sparse Representations for Voice Conversion." IEEE/ACM Transactions on Audio, Speech, and Language Processing 28 (2020): 343–354.
[16] Haneche, Houria, Bachir Boudraa, and Abdeldjalil Ouahabi. "A New Way to Enhance Speech Signal Based on Compressed Sensing." Measurement 151 (2020): 107236.
[17] Mavaddati, Samira. "A New Method for Speech Enhancement Based on Incoherent Model Learning in Wavelet Transform Domain." Signal and Data Processing 17, no. 3 (2020): 17–36.
[18] Al-Hassani, Ihsan, Oumayma Al-Dakkak, and Abdlnaser Assami. "Phonetic Segmentation Using a Wavelet-Based Speech Cepstral Features and Sparse Representation Classifier." Journal of Telecommunications and Information Technology 4 (2021): 12–22.
[19] Kwek, Lee-Chung, Alan W. C. Tan, Heng-Siong Lim, Cheah Heng Tan, and Khaled Abdulaziz Alaghbari. "Sparse Representation and Reproduction of Speech Signals in Complex Fourier Basis." International Journal of Speech Technology 25, no. 1 (2021): 211–217.
[20] Sun, Linhui, Yunyi Bu, Pingan Li, and Zihao Wu. "Single-Channel Speech Enhancement Based on Joint Constrained Dictionary Learning." EURASIP Journal on Audio, Speech, and Music Processing 2021, no. 29 (2021): 1–14.
[21] Reddy, M. Kiran, and Paavo Alku. "Exemplar-Based Sparse Representations for Detection of Parkinson’s Disease from Speech." IEEE/ACM Transactions on Audio, Speech, and Language Processing 31 (2023): 1386–1396.
[22] Xie, Yuan, Kan Xie, and Shengli Xie. "Underdetermined Blind Source Separation of Speech Mixtures Unifying Dictionary Learning and Sparse Representation." International Journal of Machine Learning and Cybernetics 12, no. 12 (2021): 3573–3583.
[23] Chen, Xiaoju, Y. Li, S. Ding, B. Tan, and Y. Jiang. "A Novel Nonlinear Dictionary Learning Algorithm Based on Nonlinear-KSVD and Nonlinear-MOD." In Lecture Notes in Computer Science, 167–179. 2022.
[24] Nejati, Mansour. "Improving the Performance of Sparse Representation Model for Image Restoration and Reconstruction." PhD diss., Isfahan University of Technology, 2016.
[25] Cai, Shuting, Shaojia Weng, Binling Luo, Daolin Hu, Simin Yu, and Shuqiong Xu. "A Dictionary-Learning Algorithm Based on Method of Optimal Directions and Approximate K-SVD." In 2016 35th Chinese Control Conference (CCC), 4793–4798. 2016.
[26] Deeba, Farah, and Kun She. "Lossless Digital Image Watermarking in Sparse Domain by Using K-Singular Value Decomposition Algorithm." IET Image Processing 14, no. 10 (2020): 2080–2087.
[27] Parsa, Javad, Mostafa Sadeghi, Massoud Babaie-Zadeh, and Christian Jutten. "Joint Low Mutual and Average Coherence Dictionary Learning." In Proceedings of the 26th European Signal Processing Conference (EUSIPCO 2018), 1739–1743. 2018.
[28] Parsa, Javad, Mostafa Sadeghi, Massoud Babaie-Zadeh, and Christian Jutten. "A New Algorithm for Dictionary Learning Based on Convex Approximation." In Proceedings of the 27th European Signal Processing Conference (EUSIPCO 2019). 2019.
مدل سازی در مهندسی
دوره 24، شماره 85
در حال تکمیل شدن
تیر 1405
صفحه 17-32

فایل ها

سابقه مقاله

  • تاریخ دریافت: 18 اردیبهشت 1403
  • تاریخ بازنگری: 09 فروردین 1404
  • تاریخ پذیرش: 03 تیر 1404

هم رسانی

ارجاع به این مقاله

آمار