Development of a speech recognition application

Abstract

The article is devoted to real-time speech recognition technology and its application in various spheres of life and activity. The main approaches, methods and factors that affect the quality and speed of speech recognition are considered. It shows how you can create a test application for speech recognition based on the Microsoft Speech API, which allows the user to dictate commands in Russian and see their text representation on the screen. The article describes the main functional requirements for the application. Ways to improve the quality of speech recognition using an adaptation model that allows you to increase the probability of recognizing certain words or phrases are considered. An example of the system interface is given.