System of automatic subtitling of video files

This paper contains a description of the process of automatic subtitling (recognition of the audio component) of video files in the process of TV production of the State Television and Radio Broadcasting Company "Volgograd-TRV". The theoretical significance of the work lies in the developed algorithms for extracting an audio track from a video file, speech recognition in an audio file, correcting recognized text, adding subtitles to a video file. The developed software module for automatic subtitling of video files uses the following technologies: the Russian language recognition model Golos ("Sber"), the Emby open source media server.