N.A. Borsuk − Ph.D. (Eng.), Associate Professor, 

«Information Systems and Networks» Department, Kaluga Branch of Bauman MSTU

E-mail: borsuk.65@yandex.ru

E.O. Deryugina − Ph.D. (Eng.), Associate Professor, 

«Information Systems and Networks» Department, Kaluga Branch of Bauman MSTU

E-mail: syvorova_eo@mail.ru

V.A. Gartman − Graduate Student, 

«Information Systems and Networks» Department, Kaluga Branch of Bauman MSTU

E-mail: vadim.gartman@yandex.ru

This article is dedicated to the development of a voice assistant for the system for visually impaired readers. This article describes how to recognize the human voice. The first step of this algorithm is to convert the sound wave into pulses that the computer can process. Further, the received signal is broken into frames. Breaking occurs «overlapped» so as not to miss information at the joints of frames. For more accurate recognition of phonemes, frames are removed from foreign noise, after which the acoustic model passes and the system with machine learning determines variants of spoken words and the context of the sentence. In addition, the article presents the development of a voice assistant program in the Python programming language using Speech Recognition libraries for human speech recognition, pyttsx3 for speech synthesis, fuzzy wuzzy for fuzzy comparison of the obtained data. To implement the voice assistant, the commands that it must execute are defined. For this purpose, a dictionary has been created, which includes the type of command, phrases that must correspond to this type of command and all types of calls to the voice assistant. In the running state, the program analyzes background noise and then removes it, to better recognize the command. The resulting speech instruction is not clearly compared with all possible options for saying the instructions from the dictionary, and then select the most appropriate one. The conditional operator then performs the action assigned to the command. Further development of the developed program will be complete reworking of the program under.NET technology and expansion of voice assistant functionality.

1. Borsuk N.A., Deyugina E.O., Gartman V.A. Razrabotka spetsializirovannyi bibliotechnoi sistemy. Elektromagnitnye volny i elektronnye sistemy. 2019. Т. 24. № 3. S. 40−44. DOI: 10.18127/j15604128-201903-08 (in Russian).
2. Borsuk N.A., Deyugina E.O., Gartman V.A. Avtomatizatsiya protsessa knigovydachi v spetsializirovannykh bibliotechnykh sustemakh. Elektromagnitnye volny i elektronnye sistemy. 2019. Т. 24. № 7. С. 30-37. DOI: 10.18127/j15604128-201907-05 (in Russian).
3. Grechkina Zh.V. Infornatsionno-bibliotechnoe obsluzhivanie lyudei s ogranichennymi vozmozhnostyami: Ucheb. posobie. Stavropol′: SKFU. 2016. 239 s. URL: http://biblioclub.ru/index.php-page=book&id=466960 (12.02.2020) (in Russian).
4. Golubenko N.B. Biblioteka XXI veka. M.; Berlin: Direkt-Media. 2019. 200 s. URL: http://biblioclub.ru/index.php-page=book&id=496774 (12.02.2020) (in Russian).
5. Nishit Patak. Isskustvennyi intellect dlya .Net: rech′, yazyk i poisk. M.: Litres. 2019. 300 s. (in Russian).
6. Raspoznavanie rechi. URL: https://habr.com/ru/post/226143/ (in Russian).
7. Myasnikova E.N. Ob′ektivnoe raspoznavanie zvykov rechi. Pod red. V.M. Vaintsa. L.: Energiya. 1967. 150 s. (Biblioteka po avtomatike. Vyp. 242). Rezhim dostupa: po podpiske. URL: http://biblioclub.ru/index.php-page=book&id=110729 (data obrashcheniya: 23.02.2020). Tekst: elektronnyi (in Russian).
8. Yakimenko A.A., Vikhman V.V. Vnedrenie biometricheskoi identifikatsii v sistemy kontrolya i upravleniya dostupom. Novosibirsk: Novosibirskii gosudarstvettyi tekhnicheskii universitet. 2016. 48 s. Rezhim dostupa: po podpiske. URL: http://biblioclub.ru/index.php-page=book&id=575127 (data obrashcheniya: 12.02.2020). Tekst: elektronnyi (in Russian).
9. Shelud′ko V.M. Yazyk programmirovaniya vysokogo urovhya Python: funktsii, struktury dannykh, dopolnitel′nye moduli. Rostov-naDonu; Taganrog: Izdatel′stvo Yuzhnogo federal,nogo universiteta. 2017. 108 s. Rezhim dostupa: po podpiske. URL: http://biblioclub.ru/index.php-page=book&id=500060 (data obrashcheniya: 17.02.2020) (in Russian). 
10. Maerkin A.V. Postroenue zaprosov i programmirovanie na SQL: Ucheb. posobie. Izd. 3-е, pererab. i dop. M.: Dialog-MIFI. 2014. 384 s. Rezhim dostupa: po podpiske. URL: http: // biblioclub.ru/index.php-page=book&id=89077 (data obrashcheniya: 17.02.2020)  (in Russian).