Device, system, and method of user authentication utilizing an optical microphone

• US 10,311,219 B2
• Filed: 06/07/2016
• Issued: 06/04/2019
• Est. Priority Date: 06/07/2016
• Status: Active Grant

• Current Assignee: Vocalzoom Systems Ltd.
• Original Assignee: Vocalzoom Systems Ltd.
• Inventors: Bakish, Tal
• Primary Examiner(s): Cervetti, David Garcia
• Application Number:
  US15/175,063
• Publication Number:
  US 20170351848A1
• Time in Patent Office: 1,092 Days
• CPC Class Codes: G06F 21/32 using biometric data, e.g. fingerprints, iris scans or voiceprints
• CPC Class Codes: G06F 21/83 input devices, e.g. keyboards, mice or controllers thereof

Abstract

Device, system, and method of user authentication utilizing an optical microphone or laser-based microphone. An optical microphone transmits an outgoing optical signal or laser beam towards a face of a human speaker; receives an incoming optical feedback that is reflected back from the face of the human speaker; performs self-mix interferometry that is based on the outgoing optical signal and the incoming reflected optical signal; and generates a user-specific feature or characteristic that uniquely characterizes said human speaker. A user authentication module operates to authenticate the user for performing a privileged or an access-controlled action, based on the user-specific characteristic that was generated, optionally in combination with one or more biometric features or authentication requirements.

Claims

1. A system comprising:

a laser-based self-mix microphone (a) to transmit an outgoing laser signal towards a face of a human speaker, (b) to receive an incoming optical feedback that is reflected back from said face of said human speaker as reflected feedback to said outgoing laser signal, (c) to perform laser self-mix interferometry of said outgoing laser signal and said incoming optical feedback, and (d) based on said laser self-mix interferometry, to generate a user-specific laser self-mix characteristic that uniquely characterizes laser self-mixing of said outgoing laser signal with said incoming optical feedback that is reflected back from said face of said human speaker, wherein said user-specific laser self-mix characteristic uniquely distinguishes between said human speaker and other human speakers;

a user authentication module to authenticate said human speaker for performing a privileged action, based on said user-specific laser self-mix characteristic that was generated by said laser self-mix interferometry of the outgoing laser signal with said the incoming optical feedback that is associated with said human speaker.

2. The system of claim 1,

wherein the user authentication module is to authenticate a user based on a comparison between:

(i) a freshly-generated user-specific laser self-mix characteristic that is freshly generated for a current user who attempts to authenticate via said laser-based self-mix microphone, and (ii) a previously-generated user-specific laser self-mix characteristic that was previously generated by an authorized user that was authenticated via said laser-based self-mix microphone.

3. The system of claim 1, comprising:

a similarity estimation unit to determine a similarity score indicating a level of similarity between (i) a freshly-generated user-specific laser self-mix characteristic that is freshly generated for a current user who attempts to authenticate via said laser-based self-mix microphone, and (ii) a previously-generated user-specific laser self-mix characteristic that was previously generated by an authorized user that was authenticated via said laser-based self-mix microphone;

wherein the user authentication module is to authenticate said current user if said similarity score is greater than a pre-defined similarity threshold.

4. The system of claim 1, comprising:

an acoustic microphone to capture acoustic output that is produced by said human speaker;

a dynamic filter generator to dynamically generate a digital filter for cleaning noises from said acoustic output that was captured by the acoustic microphone, based on output produced by said laser-based self-mix microphone, and to produce a dynamically-cleaned acoustic sample;

wherein the user authentication module is to authenticate said user for performing said privileged action, based on a combination of both;

(i) said user-specific laser self-mix characteristic that was generated from said laser self-mix interferometry of said outgoing laser signal with said incoming optical feedback, and (ii) biometric similarity between said dynamically-cleaned acoustic sample and a previously-captured acoustic sample of said human speaker.

5. The system of claim 1, comprising:

an amplitude of facial vibrations estimator, to receive output generated by said laser-based self-mix microphone, and to generate therefrom a facial vibration indicator value that indicates a level of vibrations of a facial region of said human speaker;

wherein the user authentication module comprises a facial vibrations similarity detector, to detect a level of similarity between;

(i) a freshly-generated facial vibrations indicator value, that is freshly generated for a current user who attempts to authenticate via said laser-based self-mix microphone, and (ii) a previously-generated facial vibrations indicator value that was previously generated by an authorized user that was authenticated via said laser-based self-mix microphone;

wherein said user authentication module is to authenticate said current user based on output indicating said level of similarity produced by said facial vibrations similarity detector.

6. The system of claim 1, comprising:

a speckles pattern detector, to extract, from a laser self-mix signal generated by laser self-mixing at said laser-based self-mix microphone, a unique user-specific laser-based pattern of laser-related speckles that characterize the incoming optical feedback that is reflected back from said face of said human speaker and indicates laser-based speckles detected on said human speaker;

wherein the user authentication module comprises a speckles pattern similarity detector, to detect a level of similarity between;

(i) a freshly-generated speckles pattern, that is freshly generated for a current user who attempts to authenticate via said laser-based self-mix microphone, and (ii) a previously-generated speckles pattern that was previously generated by an authorized user that was authenticated via said laser-based self-mix microphone.

7. The system of claim 1, comprising:

a speckles vibrations detector, to extract, from a laser self-mix signal generated by laser self-mixing at said laser-based self-mix microphone, a unique user-specific laser-based vibrations pattern of vibrating laser speckles that characterize the incoming optical feedback that is reflected back from said face of said human speaker and indicates vibrations of laser-based speckles on said human speaker;

wherein the user authentication module comprises a speckles pattern similarity detector, to detect a level of similarity between;

(i) a freshly-generated laser-based speckles vibrations pattern, that is freshly generated for a current user who attempts to authenticate via said laser-based self-mix microphone, and (ii) a previously-generated speckles vibrations pattern that was previously generated by an authorized user that was authenticated via said laser-based self-mix microphone.

8. The system of claim 1, comprising:

a speckles amount detector, to extract, from a laser self-mix signal generated by said laser-based self-mix microphone, a unique user-specific number indicating an amount of speckles that characterize the incoming optical feedback that is reflected back from said face of said human speaker;

wherein the user authentication module comprises a speckles pattern similarity detector, to detect a level of similarity between;

(i) a freshly-generated speckles amount number, that is freshly generated for a current user who attempts to authenticate via said laser-based self-mix microphone, and (ii) a previously-generated speckles amount number that was previously generated by an authorized user that was authenticated via said laser-based self-mix microphone.

9. The system of claim 1, further comprising:

a replay attack detector to detect that a fraudulent authentication attempt is attempted by a replay of a non-fresh previously-recorded acoustic sample, based on a difference between (I) fresh laser-based speckles noise of a freshly-captured reflected optical signal that is reflected back as feedback to said outgoing laser signal, and (II) previously-determined speckles noise that characterized a previous sample of a previously-authenticated user.

10. The system of claim 1, further comprising:

a liveness detector to detect liveness of a current user who attempts to authenticate by uttering a speech, wherein the liveness detector is to compare between;

(I) a freshly generated output of said laser-based self-mix microphone that receives fresh incoming optical feedback reflected back from said face of said current user in response to said outgoing laser signal, and (II) an expected optical feedback signal that is expected to be reflected from said current user if said current user is actually uttering said speech while being hit by said outgoing laser signal.

11. The system of claim 1, further comprising:

a liveness detector to detect liveness of a current user who attempts to authenticate by uttering a speech, wherein the liveness detector is to estimate facial skin elasticity of the human speaker based on spectral analysis of reflected optical feedback signals that are reflected back from said face of said human speaker in response to said outgoing laser signal transmitted by said laser-based self-mix microphone.

12. The system of claim 1, wherein the laser-based self-mix microphone comprises a laser beam transmitter to transmit said outgoing laser signal, and further comprises a laser self-mix interferometry unit to perform said laser self-mix interferometry of said outgoing laser signal with said reflected optical feedback, and is part of a hybrid optical-and-acoustic sensor which is further able to acquire an acoustic signal of said human speaker.

13. The system of claim 1, wherein the laser-based self-mix microphone comprises a laser beam transmitter to transmit said outgoing laser signal, and further comprises a laser self-mix interferometry unit to perform said laser self-mix interferometry of said outgoing laser signal with said reflected optical feedback, and is embedded in an electronic device selected from the group consisting of:

a smartphone, a tablet, a laptop computer.