Multi-Label Extreme Learning Machine (MLELMs) for Bangla Regional Speech Recognition
Source abstract
Abstract
Extensive research has been conducted in the past to determine age, gender, and words spoken in Bangla speech, but no work has been conducted to identify the regional language spoken by the speaker in Bangla speech. Hence, in this study, we create a dataset containing 30 h of Bangla speech of seven regional Bangla dialects with the goal of detecting synthesized Bangla speech and categorizing it. To categorize the regional language spoken by the speaker in the Bangla speech and determine its authenticity, the proposed model was created; a Stacked Convolutional Autoencoder (SCAE) and a Sequence of Multi-Label Extreme Learning machines (MLELM). SCAE creates a detailed feature map by identifying the spatial and temporal salient qualities from MFEC input data. The feature map is then sent to MLELM networks to generate soft labels and then hard labels. As aging generates physiological changes in the brain that alter the processing of aural information, the model took age class into account while generating dialect class labels, increasing classification accuracy from 85% to 95% without and with age class consideration, respectively. The classification accuracy for synthesized Bangla speech labels is 95%. The proposed methodology works well with English speaking audio sets as well.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Machine Learning and ELM · Domain Adaptation and Few-Shot Learning · MicroRNA in disease regulation
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Department of Computer Science and Engineering, Brac University, Dhaka 1212, Bangladesh
Media Laboratory, Kyung Hee University, Yong-in 17104, Korea
Department of Computer Science and Engineering, Sejeong University, Seoul 05006, Korea
Funding
G_ITRC, the Ministry of Science and ICT (MSIT), South Korea, under the Grand Information Technology Research Center Support Program · Institute for Information & Communications Technology Planning & Evaluation (IITP)
References (36)
Show source-provided references
1. Development of annotated Bangla speech corpora
2. Honnet, P.-E., Lazaridis, A., Garner, P.N., and Yamagishi, J. (2020, July 07). The siwisfrench speech synthesis database-design and recording of a high quality french database for speech synthesis. J. Idiap Tech. Rep., Available online: https://www.researchgate.net/publication/315893580_The_SIWIS_French_Speech_Synthesis_Database_-_Design_and_recording_of_a_high_quality_French_database_for_speech_synthesis.
3. Pial, T.I., Aunti, S.S., Ahmed, S., and Heickal, H. (2018, January 10–12). End-to-End Speech Synthesis for Bangla with Text Normalization. Proceedings of the 5th International Conference on Computational Science/ Intelligence and Applied Informatics (CSII), Yonago, Japan.
4. Rahman, S., Kabir, F., and Huda, M.N. (2015, January 10–12). Automatic gender identification system for Bengali speech. Proceedings of the 2nd International Conference on Electrical Information and Communication Technologies (EICT), Khulna, Bangladesh.10.1109/EICT.2015.7392013 ↗
5. Hassan, F., Khan, M.S.A., Kotwal, M.R.A., and Huda, M.N. (2012, January 18–19). Gender independent Bangla automatic speech recognition. Proceedings of the 2012 International Conference on Informatics, Electronics and Vision (ICIEV), Dhaka, Bangladesh.10.1109/ICIEV.2012.6317500 ↗
6. A study on the challenges and opportunities of speech recognition for Bengali language10.1007/s10462-021-10083-3 ↗
7. Gutkin, A., Ha, L., Jansche, M., Pipatsrisawat, K., and Sproat, R. (2016, January 23–28). TTS for Low Resource Languages: A Bangla Synthesizer. Proceedings of the 2016-10th International Conference on Language Resources and Evaluation, Portoroz, Slovenia.
8. Sadeque, F.Y., Yasar, S., and Islam, M.M. (2013, January 17–18). Bangla text to speech conversion: A syllabic unit selection approach. Proceedings of the 2013 International Conference on Informatics, Electronics and Vision (ICIEV), Dhaka, Bangladesh.10.1109/ICIEV.2013.6572593 ↗
9. Alam, F., Nath, P.K., and Khan, M. (2007). Text to speech for Bangla language using festival. Project: Bangla Language. [Processing. Thesis, Brac University Library]. Available online: http://hdl.handle.net/10361/675.
10. Muhammad, G., Alotaibi, Y.A., and Huda, M.N. (2009, January 21–23). Automatic speech recognition for Bangla digits. Proceedings of the 2009 12th International Conference on Computers and Information Technology, Dhaka, Bangladesh.10.1109/ICCIT.2009.5407267 ↗
11. Asfak-Ur-Rahman, M., Kotwal, M.R.A., Hassan, F., Ahmmed, S., and Huda, M.N. (2012, January 22–24). Gender effect cannonicalization for Bangla ASR. Proceedings of the 15th International Conference on Computer and Information Technology (ICCIT), Chittagong, Bangladesh.10.1109/ICCITechn.2012.6509701 ↗
12. Building Statistical Parametric Multi-speaker Synthesis for Bangladeshi Bangla10.1016/j.procs.2016.04.049 ↗
13. Rahut, S.K., Sharmin, R., and Tabassum, R. (2020, January 21–22). Bengali Abusive Speech Classification: A Transfer Learning Approach Using VGG-16. Proceedings of the 2020 Emerging Technology in Computing, Communication and Electronics (ETCCE), Dhaka, Bangladesh.10.1109/ETCCE51779.2020.9350919 ↗
14. Badhon, S.M., Rahaman, H., Rupon, F.R., and Abujar, S. (2021). Bengali Accent Classification from Speech Using Different Machine Learning and Deep Learning Techniques. Soft Computing Techniques and Applications, Springer.10.1007/978-981-15-7394-1_46 ↗
15. Alam, T., Khan, A., and Alam, F. (2022, March 25). Bangla Text Classification using Transformers Project: Bangla Language Processing. Available online: https://www.researchgate.net/publication/345654685_Bangla_Text_Classification_using_Transformers.
16. Bangla hate speech detection on social media using attention-based recurrent neural network
17. Trends in audio signal feature extraction methods10.1016/j.apacoust.2019.107020 ↗
18. Age-related differences in vocal responses to pitch feedback perturbations: A preliminary study10.1121/1.3273880 ↗
19. Jia, Y., Zhang, Y., Weiss, R., Wang, Q., Shen, J., Ren, F., Nguyen, P., Pang, R., and Lopez Moreno, I. (2019). Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis. arXiv.
20. Jam, M.M., and Sadjedi, H. (2009, January 2–4). Identification of hearing disorderly multi-band entropy cepstrum extraction from infant’s cry. Proceedings of the International Conference on Biomedical and Pharmaceutical Engineering, Singapore.10.1109/ICBPE.2009.5384066 ↗
21. Ribeiro, A., Matos, L.M., Pereira, P.J., Nunes, E.C., Ferreira, A.L., Cortez, P., and Pilastri, A. (2020). Deep Dense and Convolutional Auto-Encoders for Unsupervised Anomaly Detection in Machine Condition Sounds. arXiv.
22. Turchenko, V., and Luczak, A. (2017, January 21–23). Creation of a deep convolutional auto-encoder in Caffe. Proceedings of the 9th IEEE International Conference on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications (IDAACS), Bucharest, Romania.10.1109/IDAACS.2017.8095172 ↗
23. (2010, June 10). Nervana Systems/Neon, Convolutional Auto-Encoder Example Network for MNIST Data Set. Available online: https://github.com/NervanaSystems//examples/auto-encoder.py.
24. Deep convolutional auto-encoder for radar-based classification of similar aided and unaided human activities10.1109/TAES.2018.2799758 ↗
25. Guo, X., Liu, X., Zhu, E., and Yin, J. (2017). Deep Clustering with Convolutional Auto-Encoders. Lecture Notes in Computer Science, (Including Subseries Lecture Notes in Artificial Intelligence and Lecture, Notes in Bioinformatics), Springer. 10635 LNCS:373–382;.10.1007/978-3-319-70096-0_39 ↗
26. Ghasedi Dizaji, K., Herandi, A., Deng, C., Cai, W., and Huang, H. (2017, January 22–29). Deep Clustering via Joint Convolutional Auto-Encoder Embedding and Relative Entropy Minimization. Proceedings of the IEEE International Conference on Computer Vision, Venice, Italy.10.1109/ICCV.2017.612 ↗
27. Deep networks for motor control functions. Frontiers in computational neuroscience
28. Multi-label classification using a cascade of stacked auto-encoder and extreme-learning machines10.1016/j.neucom.2019.05.051 ↗
29. Tursunov, A., Choeh, J.Y., and Kwon, S. (2021). Age and Gender Recognition Using a Convolutional Neural Network with a Specially Designed Multi-Attention Module through Speech Spectrograms. J. Mdpi Sens., 21.10.3390/s21175892 ↗
30. Neural architectures for gender detection and speaker identification, Cogent Engineering10.1080/23311916.2020.1727168 ↗
31. Classification of regional and genre varieties of Chinese: A correspondence analysis approach based on comparable balanced corpora10.1017/S1351324920000121 ↗
32. Free classification of regional dialects of American English10.1016/j.wocn.2006.06.001 ↗
33. Extreme learning machine: Theory and applications10.1016/j.neucom.2005.12.126 ↗
34. Voxceleb: Large-scale speaker verification in the wild
35. Bengali Spoken Digit Classification: A Deep Learning Approach Using Convolutional Neural Network10.1016/j.procs.2020.04.148 ↗
36. MIT Deep Learning Genomics-Lecture11-PCA, t-SNE, Auto-Encoder Embedings (2020, June 20). Youtube, Manolis Kellis. Available online: https://www.youtube.com/watch?v=Qh6cAXJJxd4.