Speech semantic recognition: A control mechanism for assistive robot

摘要

Speech semantic recognition system is a system that can be applied on assistive robotic application that using a meaning of speech for people who suffers permanent disability to move around freely. In this project the user interface is provided in order to receive the speech input from the user and delivers the input of the speech from the user to the application. Speech signal is the input. From the speech signal, equivalent word can be identified. The word semantic model signifies the correlation between semantic databases that represent the meaning of a word with the registered word belonging to speech recognizer. The word that has been recognized then is converted to a matching word semantic model by using feature extraction approach. Later, the word that has been converted is stored in the database and trained. According to the feature matching method, the speech recognizer will recognize the speech signal again then the application will alerted thru comparing the word semantic model stored in the data before direct to the application interface. Since this system is successfully developed, this system also can be applied on Autonomous surface vehicle or unmanned surface vehicles (USV).