METHOD FOR SEPARATING SPEECH AND PAUSES BY ANALYZING CHARACTERISTICS OF SPECTRAL COMPONENTS OF MIXTURE OF SIGNAL AND NOISE Russian patent published in 2024 - IPC G10L25/93 

Abstract RU 2814115 C1

FIELD: computer technology for digital processing of voice information.

SUBSTANCE: spectral analysis is carried out by analysing multi-frequency periodic signals represented by digital samples using compensation of cross products. For several positions of the sliding window, based on the results of the analysis of the values of the interference parameters, threshold values are set for the powers and number of detected spectral components, as well as for the average value of the powers of the spectral components of the interference. For sliding window positions for which a signal may be present, the average number of detected components is calculated. If this number exceeds the corresponding threshold, then the signal duration is calculated. If the signal duration exceeds the minimum threshold value and does not exceed the maximum threshold value, then the average value of the powers of the spectral components of the mixture of speech signal and noise is calculated. If this value exceeds the threshold, then a signal is considered to be present in these sliding windows. Otherwise, it is decided that there is only interference in these time slots.

EFFECT: increasing the accuracy of speech and pause separation in the presence of interference with rapidly changing power.

1 cl, 3 dwg, 1 tbl

Similar patents RU2814115C1

Title Year Author Number
METHOD FOR SEPARATING SPEECH AND PAUSES BY ANALYZING CHANGES IN VALUES OF FREQUENCY AND TIME PARAMETERS OF ADDITIVE MIXTURE OF SIGNAL AND NOISE 2023
  • Zolotarev Vladimir Alekseevich
  • Karmanov Danil Sergeevich
RU2811741C1
METHOD OF DIVIDING SPEECH AND PAUSES BY VALUES OF DISPERSIONS OF AMPLITUDES OF SPECTRAL COMPONENTS 2019
  • Belogurov Vladimir Aleksandrovich
  • Zolotarev Vladimir Alekseevich
RU2723301C1
METHOD OF DETERMINING PRESENCE OF HARMONICS IN A SIGNAL, THE DURATION OF WHICH EXCEEDS A SPECIFIED VALUE 2023
  • Zolotarev Vladimir Alekseevich
RU2821216C1
METHOD FOR ISOLATING A SPEECH SIGNAL USING TIME-DOMAIN ANALYSIS OF THE SPECTRUM OF AN ADDITIVE MIXTURE OF A SIGNAL AND ACOUSTIC INTERFERENCE 2022
  • Belogurov Vladimir Aleksandrovich
  • Zolotarev Vladimir Alekseevich
RU2786547C1
METHOD FOR SPEECH EXTRACTION BY ANALYSING AMPLITUDE VALUES OF INTERFERENCE AND SIGNAL IN TWO-CHANNEL SPEECH SIGNAL PROCESSING SYSTEM 2022
  • Zolotarev Vladimir Alekseevich
  • Karmanov Danil Sergeevich
RU2807194C1
METHOD OF SEPARATION OF SPEECH AND PAUSE BY COMPARATIVE ANALYSIS OF INTERFERENCE POWER VALUES AND SIGNAL-INTERFERENCE MIXTURE 2017
  • Belogurov Vladimir Aleksandrovich
  • Zolotarev Vladimir Alekseevich
RU2668407C1
METHOD OF SEPARATING SPEECH AND SPEECH-LIKE NOISE BY ANALYZING VALUES OF ENERGY AND PHASES OF FREQUENCY COMPONENTS OF SIGNAL AND NOISE 2019
  • Belogurov Vladimir Aleksandrovich
  • Zolotarev Vladimir Alekseevich
RU2700189C1
METHOD OF SEPARATION OF SPEECH AND PAUSES BY ANALYSIS OF THE VALUES OF PHASES OF FREQUENCY COMPONENTS OF NOISE AND SIGNAL 2018
  • Belogurov Vladimir Aleksandrovich
  • Zolotarev Vladimir Alekseevich
RU2680735C1
METHOD OF SEPARATING SPEECH AND PAUSES BY ANALYZING VALUES OF INTERFERENCE CORRELATION FUNCTION AND SIGNAL AND INTERFERENCE MIXTURE 2018
  • Belogurov Vladimir Aleksandrovich
  • Zolotarev Vladimir Alekseevich
RU2691603C1
METHOD OF SPECTRAL ANALYSIS OF MULTIFREQUENCY PERIODIC SIGNALS USING COMPENSATION OF COMBINATION COMPONENTS 2019
  • Belogurov Vladimir Aleksandrovich
  • Zolotarev Vladimir Alekseevich
RU2730043C1

RU 2 814 115 C1

Authors

Zolotarev Vladimir Alekseevich

Dates

2024-02-22Published

2023-08-09Filed