technology apps in public security systems
The main scope of work will concern the development of the first three modules. In the first module, the speech signal will undergo time-frequency analysis, divided into segments of appropriate length, and assigned certain parameters. This will allow for providing adequate data for recognizing both the speaker and the speech in the next two modules.
Thanks to speech recognition, it will be possible to put text into a computer using a microphone, instead of a keyboard, and to control selected functions of the computer. The text obtained from the speech recognition module will undergo linguistic analysis in order to decrease the number of recognition errors. A semantic analysis of the text will make it possible to select a proper answer for the computer, both in the written and the spoken form because the computer will be equipped in a speech synthesizer. The key goal is to develop systems that will work in real time. The planned hierarchic structure of the system will allow for detecting and correcting errors occurring at successive stages of recognition. The module performing the syntax and semantic analysis and selecting the responses, on the basis of its results, will constitute an implementation of an artificial intelligence algorithm.
The system will be used by, among others, by the Police for the purpose of increasing the efficiency and effectiveness of the operational-reconnaissance and investigative activities.
The objective of the project is to develop a prototype system for speech analysis connected with a voice interface. The system will eventually consist of the following five modules:
Project coordinator:
The institution performing the project:
Project financed by The National Centre for Research and Development.
Get in touch with us at projekty@ppbw.pl
