Речевой сигнал и индивидуальность голоса

The monograph is devoted to the problem of automatic identification of voices. It will touch upon a range of issues related to the study of the individual characteristics of the voice, which manifests itself in the process of real human speech activity. The role of both individual phonemes and their combinations, as well as more complex semantic units of speech in the transmission of information about the individuality of the voice is discussed in detail.

Assessing the effectiveness of phonemes under conditions of their isolated pronunciation;
On the role of the phonetic quality of sounds in continuous speech;
Assessing the significance of spoken segments of speech in recognizing individuals.
On distinguishing voices under conditions of limited speech spectrum signal by spectrum and amplitude:
On relative description and naturalness;
Distinguishing voices transmitted over a path with a variable bandwidth;
Comparison of voice recognition by a natural speech signal and amplitude-limited speech.
Spectral analysis of the speech signal and characterization of the speaker's voice differences:
Spectral-temporal representation of the speech signal. Analyzers and videographers;
Some comments on the possibilities of visual examination of dynamic and contour spectrograms;
The "keyword" method in automatic voice recognition;
Features of the integrated speech spectrum and its application in automatic voice identification tasks ;
On the significance of the formant parameters of a speech signal in the characteristics of voices.
Analysis of the main temporal parameters of a speech signal:
Analysis of the microtemporal structure of a speech signal and automatic voice discrimination;
Analysis of intervals of the main tone of voice and functions of their distribution;
Evaluation of the effectiveness of the pitch parameter in distinguishing speakers;
On the nature of the macrotemporal structure of the speech signal.
Studying the amplitude characteristics of speech:
On the intensity of the speech signal and the degree of its participation in the characteristics of an individual voice;
On the nature of the distributions of the instantaneous values ​​of the amplitudes of the original speech signal and its envelope.
Chapter six. The study of the prosodic characteristics of speech in their connection with the personality of the speaker's pronunciation:
On the duration of pronunciation and the pace of speech;
On the individual nature of pauses in speech;
The role of intonation features of pronunciation in voice recognition.
Development of automatic voice identification systems:
Some issues of improving the efficiency of the description of recognition objects and the decision-making process;
Problems of identifying and verifying the identity of the speakers;
The "Integral" algorithm and its capabilities;
Device "Voice key";
Prosodika algorithm and its capabilities;
Complex automatic voice recognition system;
The problem of imitation and verification of verification systems for reliability.