Embedding speech recognition tools for custom software: Engines Overview

Abstract

Different solutions and tools for speech recognition are now available. Nevertheless, implementation of natural language processing still remains a current problem. Developing any custom software with a good style of UI/UX requires the integration of speech recognition. Evidently, the most common solution is to use some engine as an embedded standard tool. Here in the paper we are presenting an overview and an analysis of some popular speech recognition engines: Google Speech Recognition API, Microsoft Speech API, Yandex Speech Kit and Julius. These speech recognition tools are a readyto- serve and suitable to supplement your own software with a reliable voice command detection or voice control feature. The results of our analysis comes from an experiment of voice recognition using these tools as an embedded component in a custom software.

Description

Keywords

speech recognition, speech engine, API, voice command detection, voice control, Google, Microsoft, Yandex, Julius, overview and analysis

Citation

Dovbysh A. Embedding speech recognition tools for custom software: Engines Overview / Arthur Dovbysh, Vladyslav Alieksieiev // Computational linguistics and intelligent systems, 25-27 June 2018. — Lviv : Lviv Polytechnic National University, 2018. — Vol 2 : Workshop. — P. 114–121. — (Part 2. Workshop conference tracks. Section I. Computational Linguistics).