Využití virtuální instrumentace pro zpracování řečových signálů v oblasti SMART technologií a Průmyslu 4.0

Abstract

This thesis deals with automatic speech recognition in Industry 4.0 and SMART technology for subsequent testing of selected filtration methods. Firstly, the thesis deals with the search of voice control in the Industry 4.0 and SMART technologies. It also deals with methods of automatic speech recognition and filtering methods. Primarily, this work focuses on voice control and interference filtering using adaptive LMS and Independent Component Analysis (ICA). In this work, a software application was created to create a jam recording database. Based on these recordings, three visualizations were made to test selected methods. Recognition success is evaluated by state recognized / not recognized when each command was 100x pronounced.