| Abstract [eng] |
One of the first and longest-used forms of communication between people is language. With the rapid development of technologies, humans began to use language not only for communication with each other, but they also began to communicate with technology. The main purpose of this is that devices would be controlled not by physical work, but by spoken voice commands. The devices, which are controlled by voice, are aimed at people having visual or mobility disabilities. Until now, developing voice-recognition technology is designed to manage a single system, but it causes inconvenience for the consumer who wants to manage several different systems at the same time. One of the main purposes of this thesis is the creation of a multifunctional voice controlled system, for voice control of six different systems. After a detailed analysis of the literature, covering speech recognition technologies, the identification of system operation principle and methods used in speech recognition modeling. Research of speech recognition was accomplished by using other language recognition system and Lithuanian recognition system. The studies of spoken voice commands are conducted using the hidden Markov models applicable methods, non-native identifiers of speech recognition is used for other language recognition, WEKA data analysis system, the classificators used in Auto-WEKA and others. A Multifuncional voice controlled system is handed in this thesis, recognising voice commands witch are preprogramed to control specific devices. The program is written in C # programming language. The voice command recognition accuracy tests are carried then out. The obtained results were analyzed and conclusions were given. |
