Search academic texts

Information × Registration Number 0214U006137, 0112U000565 , R & D reports Title Development of program libraries of automatic speaker recognition on the basis of robust identification features popup.stage_title Head Shelepov Vladislav Yurievich, Registration Date 12-02-2014 Organization Institute of Artificial Intelligence Problems popup.description2 A digital speech signal is the object of current research. The project aims to create methodic, hardware and software products for solving speaker identification task using robust identification features. The methods of parameterization and signal recognition are used. They are well-known and widely used in modern systems of voice biometrics. Were created: - The software libraries for an automated speaker identification; - A new hybrid approach for calculating the trajectories of formants and melodic contour; - Improved signal pre-processing technology in order to determine the boundaries of speech, reduce noise and to determine the suitability to the identification of a phonogram; - Method of robust attribute descriptions forming in the tasks of group and direct speaker identification; - Improved technology of intonation phrases type determination; - Improved speaker identification technology. Industries: - In the manufacturing sector during phonoscope research on the orders or decrees of law enforcement and government agencies; - Hardware and software systems: check access rights to different systems, speech biometric identification, speech synthesis , audio indexing on the basis of changes in speaker; - The development of teaching and hardware and software determine emotional state in her speech. Product Description popup.authors Єрмоленко Тетяна Володимирівна Дорохіна Галина Володимирівна Клименко Микита Сергійович Ніценко Артем Володимирович Шелєпов Владислав Юрійович popup.nrat_date 2020-04-02 Close