Recognizing speech in voice messages
چکیده
The level of development information technology makes it possible to use speech recognition technologies in a wide range human life and activities. It is very convenient the voice interface: search for necessary documents, dialing phone number, managing IOT devices, navigation, simple text dictation. Since natural language interface provides an additional convenience person when typing, sending messages has become common among users. In this case, are audio files. But not always available recipient listen such messages. This problem can be solved with help automatic system (ASR). article describes stages elements process processing by signal. Modern problems choosing them indicated. (ASR) systems understand fully spontaneous that natural, memorized, contains signs stuttering or even minor errors. At same time, they still too expensive develop from scratch. So companies faced choice between using cloud API ASR developed tech giants open source solutions. analysis latest research publications on data considered. A software solution conversion into proposed. signal delivery proposed made as chat bot messenger. presents main components system, algorithm bot, modern development, implementation configuration messenger