Corresponding author: Korsun Oleg N., D. Sc., Professor, State Research Institute of Aviation Systems, State Scientific Center of the Russian Federation, Moscow, 125319, Russian Federation, e-mail: marmotto@rambler.ru

Received on April 23, 2015
Accepted on May 15, 2015

The topic of the article is the problem of applicability of the developed robust algorithms for an automatic recognition of the speech commands to the other languages in order to improve the interface of the onboard equipment of modern aircraft and other technical systems. Resistance to acoustic noise is one of the most important requirements to the on-board equipment of the voice control system. This paper describes the algorithms of parameterization of the speech signal, a method for formation of a pattern and algorithm for recognition of the individual words by comparison with the standards of words. Also, a noise resistance algorithm is described, which is based on a linear regression and application of an additional microphone, and intended to improve the acoustic noise resistance of the automatic speech command recognition. In this algorithm the mix of portraits of the recognized words and noises are compared simultaneously with the patterns and the signals from an additional microphone. The additional microphone is positioned at a distance of 0,3—0,7 m from a speaker in order to fix predominantly the acoustic noises in the cockpit. The report presents the results of several experiments for the Russian, Kazakh and Tajik languages, showing a high probability of recognition of the commands for the on-board equipment control in the conditions of additional noises with the level of 80—90 dB, which corresponds to the maximal admissible noise values in the cockpit of a modern long-range aircraft. The results also prove that the described algorithms can work well and do not require any changes for recognition of words in other languages.

Keywords: automatic speech recognition; noise-resistant algorithms of speech recognition; voice control of the aircraft equipment, algorithm invariant to languages

Acknowledgements: This work was supported by a grant from the Russian Foundation for Basic Research, project No. 12-08-00670-a.

For citation:
Korsun O. N., Gabdrakhmanov A. Sh., Mihajlov E. I., Nahaev M. Z., Tulekbaeva A. K. Algorithm for an Automatic Recognition of the Speech Commands, Invariant to Languages, Mekhatronika, Avtomatizatsiya, Upravlenie, 2015, vol. 16, no. 9, pp. 599—604.
DOI: 10.17587/mau.16.599-604

To the contents