O. N. Faizulaieva, I. S. Nevlyudov

Read the full article 


 T Scientific task for improving the signal-to-noise ratio for user’s voice signal in computer systems and networks during the process of user’s voice authentication is considered. The object of study is the process of input and extraction of the voice signal of authentication system user in computer systems and networks. Methods and means for input and extraction of the voice signal on the background of external interference signals are investigated. Ways for quality improving of the user’s voice signal in systems of voice authentication are investigated experimentally. Firmware means for experimental unit of input and extraction of the user’s voice signal against external interference influence are considered. As modern computer means, including mobile, have two-channel audio card, two microphones are used in the voice signal input. The distance between sonic-wave sensors is 20 mm and it provides forming one direction pattern lobe of microphone array in a desired area of voice signal registration (from 100 Hz to 8 kHz). According to the results of experimental studies, the usage of directional properties of the proposed microphone array and space-time processing of the recorded signals with implementation of constant and adaptive weighting factors has made it possible to reduce considerably the influence of interference signals. The results of firmware experimental studies for input and extraction of the user’s voice signal against external interference influence are shown. The proposed solutions will give the possibility to improve the value of the signal/noise ratio of the useful signals recorded up to 20 dB under the influence of external interference signals in the frequency range from 4 to 8 kHz. The results may be useful to specialists working in the field of voice recognition and speaker discrimination.

Keywords:  authentication, array, direction pattern, voice signal, quadrature processing, microphone

Acknowledgements. Работа выполнена при поддержке Министерства образования и науки Российской Федерации и Совета по грантам Президента Российской Федерации (грант МД-1072.2013.9) и частично при государственной финансовой поддержке ведущих университетов Российской Федерации (субсидия 074-U01).


1. He Y., Kavukcuoglu K., Wang Y., Szlam A., Qi Y. Unsupervised Feature Learning by Deep Sparse Coding [Электронный ресурс]. 2013. Режим доступа: http://arxiv.org/pdf/1312.5783v1, свободный. Яз. англ. (дата обращения 03.07.2014).

2. Arnold L., Rebecchi S., Chevallier S., Paugam-Moisy H. An introduction to deep learning // Proc. 19th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2011). Bruges, Belgium, 2011. P. 477–488.

3. Ciresan D.C., Meier U., Masci J., Schmidhuber J. Multi-column deep neural network for traffic sign classification // Neural Networks. 2012. V. 32. P. 333–338.

4. Mnih V., Kavukcuoglu K., Silver D., Graves A., Antonoglou I., Wierstra D., Riedmiller M. Playing Atari with Deep Reinforcement Learning [Электронный ресурс]. 2013. Режим доступа: http://arxiv.org/pdf/1312.5602v1.pdf, свободный. Яз. англ. (дата обращения 03.07.2014).

5. Le Roux N., Bengio Y. Representational power of restricted boltzmann machines and deep belief networks // Neural Computation. 2008. V. 20. N 6. P. 1631–1649.

6. Gregor K., Mnih A., Wierstra D., Blundell C., Wiersta D. Deep Autoregressive Networks [Электронный ресурс]. 2013. Режим доступа: http://arxiv.org/pdf/1310.8499v2, свободный. Яз. англ. (дата обращения 03.07.2014).

7. Tenenbaum J.B., Kemp C., Griffiths T.L., Goodman N.D. How to grow a mind: statistics, structure, and abstraction // Science. 2011. V. 331. N 6022. P. 1279–1285.

8. Szegedy Ch., Zaremba W., Sutskever I., Bruna J., Erhan D., Goodfellow I., Fergus R. Intriguing properties of neural networks [Электронный ресурс]. 2014. Режим доступа: http://arxiv.org/pdf/1312.6199v4, свобод- ный. Яз. англ. (дата обращения 03.07.2014).

9. Bengio Y., Lamblin P., Popovici D., Larochelle H. Greedy layer-wise training of deep networks // Advances in Neural Information Processing Systems. 2007. V. 19. P. 153–160.

10. Hinton G.E., Osindero S., Teh Y.-W. A fast learning algorithm for deep belief nets // Neural Computation. 2006. V. 18. N 7. P. 1527–1554.

11. Ranzato M.A., Poultney Ch., Chopra S., LeCun Y. Efficient learning of sparse representations with an energy-based model // Advances in Neural Information Processing Systems. 2007. V. 19. P. 1137–1144.

12. Ciresan D.C., Meier U., Gambardella L.M., Schmidhuber J. Deep Big Simple Neural Nets Excel on Handwritten Digit Recognition [Электронный ресурс]. 2010. Режим доступа: http://arxiv.org/pdf/1003.0358, свободный. Яз. англ. (дата обращения 03.07.2014).

13. Царев Ф.Н. Совместное применение генетического программирования, конечных автоматов и искус- ственных нейронных сетей для построения системы управления беспилотным летательным аппара- том // Научно-технический вестник СПбГУ ИТМО. 2008. № 8 (53). С. 42–60.

14. Бондаренко И.Б., Гатчин Ю.А., Гераничев В.Н. Синтез оптимальных искусственных нейронных сетей с помощью модифицированного генетического алгоритма // Научно-технический вестник информа- ционных технологий, механики и оптики. 2012. № 2 (78). С. 51–55.

15. Vincent P., Larochelle H., Bengio Y., Manzagol P.-A. Extracting and composing robust features with denoising autoencoders // Proc. 25th International Conference on Machine Learning. Helsinki, Finland, 2008. P. 1096–1103. 16. LeCun Y., Cortes C., Burges C.J.C. The MNIST Database of handwritten digits [Электронный ресурс]

Copyright 2001-2017 ©
Scientific and Technical Journal
of Information Technologies, Mechanics and Optics.
All rights reserved.