Control Systems and Computers

https://doi.org/10.15407/csc.2021.04.028

Control Systems and Computers, 2021, Issue 4 (294), pp. 28-34.

UDC 004.032.26

A.V. Manokhin, student, Computer Systems Software Department of the Applied Mathematics Faculty  National Technical University of Ukraine “Igor Sikorsky Kyiv Politechnic Institute”, Peremohy Ave 37, Kyiv, Ukraine, 

N.A. Rybachok, PhD, Senior Lecturer, Computer Systems Software Department of the Applied Mathematics Faculty  National Technical University of Ukraine “Igor Sikorsky Kyiv Politechnic Institute”, Peremohy Ave 37, Kyiv, Ukraine, E-mail: rybachok@pzks.fpm.kpi.ua

ENGLISH ACCENT RECOGNITION USING DEEP MACHINE LEARNING

The article highlights aspects of the use of deep machine learning to recognize the accents of the English language. The software has been developed to determine the percentage of how close audio recordings are to each of 8 most common English accents. A convolutional neural network consisting of 2 convolutional layers, 1 max pooling layer, and 2 dense layers was trained across 2 epochs on a set of 5,516 audio recordings taken from the English Multi-speaker Corpus for Voice Cloning resource. The forecasting accuracy of 89.07% was achieved on the test data presented by 11 thousand MFCC matrices with a dimension of 50×87.

Download full text! (In English).

Keywords: neural network, deep machine learning, accent recognition, MFCC, CNN.

1. The most spoken languages worldwide in 2021 [Electronic resource] Access mode: https://www.statista.com/statistics/266808/the-most-spoken-languages-worldwide/ Date accessed: May 18, 2021.
2. Sheng LMA, Edmund MWX Deep learning approach to accent classification. Stanford University Project Report 2017.Access mode: http://cs229.stanford.edu/proj2017/final-reports/5244230.pdf Date accessed: May 18, 2021.
3. Upadhyay R., 2017. Accent Classification Using Deep Belief Network, University of Mumbai. [Accessed 18 May 2021].
   https://doi.org/10.1109/ICSC.2018.00053
4. S.Hernandez, V.Bulitko, S.Carleton, A. Ensslin, T. Goo-rimoorthee Deep learning for classification of speech accents invideo games. inAIIDE Workshops, 2018.Access mode: http://ceur-ws.org/Vol-2282/EXAG_114.pdf. Date accessed: May 18, 2021.
5. S.Duduka, H.Jain, V.Jain, H.Prabhu, P.Chawan Accent Classification using Machine Learning. International Research Journal of Engineering and Technology (IRJET). Vol .7 (issue 11). – 2020. – P. 638-641.Access mode:https://www.irjet.net/archives/V7/i11/IRJET-V7I11105.pdf. Date accessed: May 18, 2021.
6. Telegram Revenue and Usage Statistics (2020). Access mode:https://www.businessofapps.com/data/telegram-statistics/. Date accessed: May 18, 2021.
7. Telegram vs WhatsApp: The Basics. Access mode:https://www.uctoday.com/collaboration/telegram-vs-whatsapp-the-basics/.Date accessed: May 18, 2021.
8. Mel Frequency Cepstral Coefficient (MFCC) tutorial Access mode: http://practicalcryptography.com/miscellaneous/machine-learning/guide-mel-frequency-cepstral-coefficients-mfccs/Date accessed: May 18, 2021.
9. S.Duduka, H.Jain, V.Jain, H.Prabhu, P.Chawan A Neural Network Approach to Accent Classification. – International Research Journal of Engineering and Technology (IRJET). Vol.8 (issue 03). – 2021. – P. 1175-1177.Access mode https://www.irjet.net/archives/V8/i3/IRJET-V8I3359.pdf. Date accessed: May 18, 2021.

Received 20.07.2021