Patents by Inventor Aleksandr Alekseevich ALENIN

Aleksandr Alekseevich ALENIN has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).

  • Methods and devices for detecting a spoofing attack
    Patent number: 11611581
    Abstract: Methods and devices for detecting a spoofing attack are provided. The method includes receiving a speech signal; applying a Fast Fourier Transform (FFT) to the speech signal to extract FFT features therefrom and feeding the extracted FFT features to a pre-trained deep neural network to assign a first confidence score to the speech signal; applying a Discrete cosine transform (DCT) to the speech signal to extract DCT features therefrom and feeding the extracted DCT features to a pre-trained deep neural network to assign a second confidence score to the speech signal; applying a pre-trained deep convolutional network (DCN) based on an end-to-end architecture to the speech signal to assign a third confidence score to the speech signal; computing a total confidence score based on the assigned confidence scores; and comparing the computed total confidence score to a predefined threshold to detect whether the received speech signal is spoofed.
    Type: Grant
    Filed: August 26, 2020
    Date of Patent: March 21, 2023
    Assignee: ID R&D, Inc.
    Inventors: Konstantin Konstantinovich Simonchik, Anton Andreevich Pimenov, Aleksandr Alekseevich Alenin
  • METHODS AND DEVICES FOR DETECTING A SPOOFING ATTACK
    Publication number: 20220070207
    Abstract: Methods and devices for detecting a spoofing attack are provided. The method includes receiving a speech signal; applying a Fast Fourier Transform (FFT) to the speech signal to extract FFT features therefrom and feeding the extracted FFT features to a pre-trained deep neural network to assign a first confidence score to the speech signal; applying a Discrete cosine transform (DCT) to the speech signal to extract DCT features therefrom and feeding the extracted DCT features to a pre-trained deep neural network to assign a second confidence score to the speech signal; applying a pre-trained deep convolutional network (DCN) based on an end-to-end architecture to the speech signal to assign a third confidence score to the speech signal; computing a total confidence score based on the assigned confidence scores; and comparing the computed total confidence score to a predefined threshold to detect whether the received speech signal is spoofed.
    Type: Application
    Filed: August 26, 2020
    Publication date: March 3, 2022
    Applicant: ID R&D, Inc.
    Inventors: Konstantin Konstantinovich SIMONCHIK, Anton Andreevich PIMENOV, Aleksandr Alekseevich ALENIN