KR-102964251-B1 - METHOD AND APPARATUS FOR VOICE AUTHENTICATION

Abstract

The present invention relates to an apparatus and method for performing voice authentication using the voice of a subject. More specifically, the present invention relates to a voice authentication technology comprising storing a registered voice fingerprint, analyzing the frequency of a voice signal input in real time to obtain a real-time voice fingerprint, comparing the obtained real-time voice fingerprint with the registered voice fingerprint to calculate a degree of similarity, and determining whether the calculated degree of similarity exceeds a predetermined threshold, wherein the control unit performs authentication for the input voice signal based on the determination result.

Inventors

• 황정환
• 권우철

Assignees

• 엠텍비젼 주식회사

Dates

Publication Date

20260512

Application Date

20240711

Claims (14)

1. Step of the storage unit storing the registered voice fingerprint; A step in which a frequency analysis unit analyzes the frequency of a voice signal input in real time to acquire a real-time voice fingerprint; A step in which a comparison unit compares the real-time voice fingerprint obtained above with the registered voice fingerprint to calculate the degree of similarity; A step in which a control unit determines whether the calculated degree of similarity exceeds a predetermined threshold; and The above control unit includes a step of performing authentication for the input voice signal based on the above judgment result, The above-mentioned registered and real-time voice fingerprints are voiceprint tables acquired while the user speaks for a preset period of time, and The above-mentioned glottal table includes the amplitude intensity for each of a plurality of frequency bands of a voice signal at unit time intervals, Voiceprint authentication method.
2. In Article 1, The above real-time voice fingerprint and the above registered voice fingerprint are characterized by being voices in which the subject speaks a preset phrase. Voiceprint authentication method.
3. In claim 1, the step of acquiring the real-time voice fingerprint is, A step of adjusting the sync of the above voice signal; and A step comprising obtaining a real-time voice fingerprint by analyzing the frequency of the above-mentioned adjusted voice signal, Voiceprint authentication method.
4. In claim 3, the step of adjusting the sink is, A step of determining the start and end points of the above voice signal; and A step comprising performing cropping on the voice signal to retain only from the starting point to the ending point, Voiceprint authentication method.
5. In claim 4, the step of adjusting the sink is, A step of determining a real-time length, which is the length of the cropped voice signal; and A step comprising adjusting the length of the cropped voice signal so that the real-time length matches the registered length, Voiceprint authentication method.
6. delete
7. In claim 1, the amplitude intensity is, not an absolute value, but a relative value Voiceprint authentication method.
8. A storage unit for storing registered voice fingerprints; A frequency analysis unit that analyzes the frequency of a voice signal input in real time to acquire a real-time voice fingerprint; A comparison unit that calculates the degree of similarity by comparing the real-time voice fingerprint obtained above with the registered voice fingerprint; and It includes a control unit that determines whether the calculated degree of similarity exceeds a predetermined threshold, The control unit performs authentication for the input voice signal based on the judgment result, The above-mentioned registered and real-time voice fingerprints are voiceprint tables acquired while the user speaks for a preset period of time, and The above-mentioned glottal table includes the amplitude intensity for each of a plurality of frequency bands of a voice signal at unit time intervals, Voiceprint authentication device.
9. In Article 8, The above real-time voice fingerprint and the above registered voice fingerprint are characterized by being voices in which the subject speaks a preset phrase. Voiceprint authentication device.
10. In claim 8, the control unit, in acquiring a real-time voice fingerprint, Adjusting the sync of the above voice signal, Controlling the frequency analysis unit to acquire a real-time voice fingerprint by analyzing the frequency of the above-mentioned adjusted voice signal, Voiceprint authentication device.
11. In claim 10, the control unit, in adjusting the sink, Determining the start and end points of the above voice signal, Performing cropping on the voice signal to retain only the portion from the starting point to the ending point, Voiceprint authentication device.
12. In claim 11, the control unit, in adjusting the sink, Determining the real-time length, which is the length of the cropped voice signal above, and Adjusting the length of the cropped voice signal so that the real-time length matches the registered length, Voiceprint authentication device.
13. delete
14. In claim 8, the amplitude is, not an absolute value, but a relative value Voiceprint authentication device.

Description

Method and Apparatus for Voice Authentication The present invention relates to a voice authentication method and apparatus, and more specifically, to a technology for performing voice authentication by analyzing the amplitude intensity of an input voice by frequency. Existing biometric methods include iris recognition, fingerprint recognition, and facial recognition. Fingerprint recognition is a contact-based method and is vulnerable to infection by viruses and bacteria. Additionally, this contact method has the disadvantage of causing inconvenience during use due to factors such as interference from other contaminants. Non-contact methods include those utilizing camera sensors, such as iris recognition, vein authentication, and facial recognition. However, these methods suffer from a lack of accuracy in situations requiring separate lighting, such as at night, or in enclosed spaces where light is blocked. Among contactless methods, there is a voiceprint authentication method utilizing a microphone. Existing voiceprint authentication methods adopt a process where a terminal recognizing the subject's voice transmits the data to a server via a network connection, and the result returns to the terminal after passing through a specific function within the server. This process inevitably introduces latency, and there are issues where accuracy is limited by network connectivity. Accordingly, there is a need for research on methods to perform voice authentication with high accuracy directly on the terminal itself, without going through a server. FIG. 1 illustrates a block diagram of a voice authentication device (100) according to an embodiment of the present invention. FIG. 2 is a diagram illustrating a conceptual diagram of sample unit storage according to an embodiment of the present invention. FIG. 3 is a diagram illustrating a conceptual diagram of a buffer section (102) according to an embodiment of the present invention. FIG. 4 is a diagram illustrating the operation of a frequency analysis unit (103) according to an embodiment of the present invention. FIG. 5 is a diagram illustrating the concept of covering a plurality of natural frequencies with a first frequency block (103-1), which is a single block, according to an embodiment of the present invention. FIG. 6 is a diagram illustrating a hardware block diagram of each frequency block (103-1 to 103-n) according to an embodiment of the present invention. FIG. 7 is a block diagram illustrating a Green function generator (601) according to an embodiment of the present invention. FIG. 8 illustrates an example stored in a frequency memory (701) according to an embodiment of the present invention. FIG. 9 is a diagram illustrating the sequence in which a voice authentication device (100) according to an embodiment of the present invention performs sequential amplitude calculations. FIG. 10 is a diagram illustrating the concept of a voice authentication device (100) according to an embodiment of the present invention performing sequential amplitude calculations using a plurality of frequency blocks (103-1 to 103-n). FIG. 11 illustrates an example of a voiceline table obtained over time according to a first embodiment of the present invention. FIG. 12 illustrates an example of a voiceline table obtained over time according to a second embodiment of the present invention. FIG. 13 illustrates an embodiment of storing a registered voice fingerprint, which is a specific user's voice fingerprint, during a registration process according to an embodiment of the present invention. FIG. 14 illustrates a flowchart of user identification in an authentication process according to an embodiment of the present invention. FIG. 15 is a diagram illustrating the concept of a comparison unit (106) comparing a registered voice fingerprint (1501) and a real-time voice fingerprint (1502) according to an embodiment of the present invention. FIG. 16 illustrates an embodiment of aligning the sync through a progressive bar (1601) output through a display according to an embodiment of the present invention. FIG. 17 is a diagram illustrating the sequence of a sink adjustment process according to an embodiment of the present invention. Hereinafter, embodiments disclosed in this specification will be described in detail with reference to the attached drawings. Identical or similar components regardless of drawing symbols will be assigned the same reference number, and redundant descriptions thereof will be omitted. The suffixes "module" and "part" used for components in the following description are assigned or used interchangeably solely for the ease of drafting the specification and do not inherently possess distinct meanings or roles. Furthermore, in describing embodiments disclosed in this specification, if it is determined that a detailed description of related prior art could obscure the essence of the embodiments disclosed in this specification, such detailed description will be omitte