Optimalno prepoznavanje i lokalizacija izvora zvuka primenom metoda veštačke inteligencije
Kovandžić, Marko 1974-
Nikolić, Vlastimir 1954-
Antić, Dragan 1963-
Ćojbašić, Žarko 1968-
Simonović, Miloš 1973-
Ćirić, Ivan 1980-
Biografija autora: list 165;Bibliografija: listovi 154-164. Datum odbrane: 14.02.2020. Automatic control and robotics
Тhe subject of the thesis is sound source recognition and sound source localization, in real conditions, using artificial intelligence algorithms. The main goal is optimal procedure for sound source observation using artificial neural networks for signal procesing, because of their extreme procesing speed. It has to provide implementation of hybrid system capable to recognize and locate sound source in the presence of disturbances. For the training and the testing of neural networks two sets of data are provided, from two different experiments, and to increase robustness genetic algorithm is applied. The results of the investigation will contribute the existing body of acoustic observation knowledge.
This work is licensed under a
CC BY 2.0 AT - Creative Commons Attribution 2.0 Austria License.
CC BY 2.0 AT