Improvement of the learning process of the automated speaker recognition system for critical use with HMM-DNN component

摘要

The article presents the results of the adaptation of the hybrid HMM-DNN speech synthesis model for use in automatedspeaker recognition system for critical use (ASRSCU). In particular, the process of learning the HMM-DNN speechsynthesis model with the estimation of the difference between the posterior probability distributions of all HMM statesand the actual a posteriori probability distribution, calculated by DNN, and the use of semantic information in thespeaker recognition process, has been improved. The features that are observed in the sequence of frames to which theinput phonogram is divided describe this information. The obtained results allowed improving the efficiency of the textdependentspeaker recognition when using ASRSCU in a noisy acoustic environment. The article formulated measuresfor the structural integration of the HMM-DNN component in ASRSCU and describes the practical aspects of thisprocess. In particular, the choice of the type and the method of normalization of the vectors of basic informative featuresat the frame level was substantiated, the number of HMM states and GMM parameters were determined depending onthe parameters of the chosen formation model, and the procedure for interpreting the recognition results was described.The paper formulates measures to optimize the learning process of the ASRSCU with the HMM-DNN component,which will be exploited in noisy environments.