Cybernetics And Systems Analysis logo
Editorial Board Announcements Abstracts Authors Archive
Cybernetics And Systems Analysis
International Theoretical Science Journal
-->

UDC 621.317+681.849
V.I. Solovyov1, O.V. Rybalskiy2, V.V. Zhuravel3, N.V. Semenova4


1 Volodymyr Dahl East Ukrainian National University, Sievierodonetsk, Ukraine

edemsvi@gmail.com

2 National Academy of Internal Affairs, Kyiv, Ukraine

rov_1946@ukr.net

3 Kyiv Expert and Forensic Center of the Ministry
of Internal Affairs of Ukraine, Kyiv, Ukraine

fonoscopia@ukr.net

4 V.M. Glushkov Institute of Cybernetics, National Academy
of Sciences of Ukraine, Kyiv, Ukraine

nvsemenova@meta.ua

ANALYZING THE MODELS OF SPEECH RECOGNITION ON THE BASIS OF NEURAL
NETWORKS OF DEEP LEARNING FOR EXAMINATION OF DIGITAL PHONOGRAMS

Abstract. The authors analyze the models based on deep learning neural networks, on the basis of the general approach to pauses and speech signals as different types of voice information fixed in a phonogram, different in some characteristics. It is shown that such an approach allows generating the learning databases with the use of the general for pauses and signals of speech methods of preliminary processing of information. This provides a high level of unification of network learning methods intended for solution of various examination problems.

Keywords: digital audio recording devices, learning database, deep learning neural network, digital treatment of phonograms, digital phonogram, examination.



FULL TEXT

REFERENCES

  1. Solovyov V.I., Rybalskiy O.V., Zhuravel V.V. Verification of fundamental fitness of neuron networks of the deep educating for the construction of the system of exposure of editing of digital phonograms. Cybernetics and Systems Analysis. 2020. Vol. 56. N 2. P. 326–330. https://doi.org//10.1007/ s10559-020-00249-2.

  2. Rybalskiy O.V., Solovyov V.I., Chernyavsky S.S., Zhuravel V.V. Features of modern probabilistic technologies of forensic examination. Pravo i pravookhorona. 2019. N 4. P. 212–215. https://doi.org/10.36486np.2019.4.

  3. Rybalskiy O.V., Solovyov V.I., Zhuravel V.V. Systems of instrumentation for the examination of audio and video recording in Ukraine. Vestn. Polotsk. state un-ty. Ser. C. 2018. N 4. P. 15–19.

  4. Rybalskiy O.V., Zharikov Yu.F. Modern methods of checking the authenticity of magnetic phonograms in forensic acoustic examination [in Russian]. K .: NAVSU, 2003. 300 p.

  5. Solovyov V.I., Rybalskiy O.V., Zhuravel V.V. Method of exposure of signs of the digital editing in phonograms with the use of neuron networks of the deep learning. Journal of Automation and Information Sciences. 2020. Vol. 52, N 1. P. 22–28. https://doi.org/10.1615/JAutomatInfScien.v52.i1.30.

  6. Rybalsky O.V., Solovyov V.I., Zhuravel V.V. Automatic segmentation of phonograms by speech pauses. Suchasna spetsialna technika. 2018. N 1. P. 58–64.

  7. Semenova N.V., Kolechkina L.N., Nagornaya A.N. Vector optimization problems with linear criteria on an indistinctly defined combinatorial set of alternatives. Kibernetika i sistemnyj analiz. 2011. N 2. P. 88–99.

  8. Malla S. Wavelets in Signal Processing [Russian translation]. Moscow: Mir, 2005. 670 p.

  9. Sapozhkov M.A. Electroacoustics [in Russian]. Moscow: Svyaz', 1978. 272 p.

  10. Alexandrova Yu.I. Psychophysiology [in Russian]. Moscow; SPb .: Nauka, 2006. 463 p.




© 2021 Kibernetika.org. All rights reserved.