Abstract
The recent technological advancements allow to develop speech to machine interfaces. These systems can be employed in a variety of potential applications like security, smart homes, centers for people with disabilities, etc. In this context a speech to machine interface is devised. The proposed system is based on the principle of frequency domain speech recognition. The spectral analysis of speech is performed in order to extract its features. Later on these extracted features could be employed to perform actions like issuing commands for actuations, granting access to the secure services, dialing with voice, banking via telephone, accessing confidential databases, etc. A simplified system prototype is designed and developed. It accepts commands in the form of speech, extract speech signal features and employ these features for piloting the actuators. The actuators are piloted according to the speaker desire. The developed system functionality is verified. Results show a proper system operation.