Systems and methods for secure authentication based on machine learning techniques
A system described herein may provide a technique for the use of machine learning techniques to perform authentication, such as biometrics-based user authentication. For example, user biometric information (e.g., facial features, fingerprints, voice, etc.) of a user may be used to train a machine learning model, in addition to a noise vector. A representation of the biometric information (e.g., an image file including a picture of the user's face, an encoded file with vectors or other representation of the user's fingerprint, a sound file including the user's voice, etc.) may be iteratively transformed until the transformed biometric information matches the noise vector, and the machine learning model may be trained based on the set of transformations that ultimately yield the noise vector, when given the biometric information.
1. A device, comprising:
one or more processors configured to:
receive a noise vector;
receive reference data;
select one or more transformations to apply to the reference data;
apply the selected one or more transformations to the reference data;
compare the reference data, with the one or more applied transformations, to the noise vector;
iteratively repeat the selecting, applying, and comparing, until the reference data, with a particular set of applied transformations, matches the noise vector within a first threshold measure of similarity;
receive test data;
apply the particular set of applied transformations to the test data; and
determine whether the test data matches the reference data based on whether the test data, with the particular set of applied transformations, matches the noise vector within a second measure of similarity.
2. The device of claim 1 , wherein the determining whether the test data matches the reference data is performed without comparing the test data to the reference data.
3. The device of claim 1 , wherein the test data includes biometric information received from a biometric authentication component of a User Equipment (“UE”).
4. The device of claim 3 , wherein the one or more processors are further configured to:
determine whether to authenticate a user, associated with the biometric information, based on whether the test data matches the reference data.
5. The device of claim 3 , wherein the biometric information includes at least one of:
image data,
fingerprint data, or
audio data.
6. The device of claim 1 , wherein the iteratively repeated selecting and applying are performed by a generator component of a Generative Adversarial Network (“GAN”), and wherein the iteratively repeated comparing is performed by a discriminator component of the GAN.
7. The device of claim 1 , wherein the noise vector is different from one or more other noise vectors used to determine whether test data matches other reference data.
8. A non-transitory computer-readable medium, storing a plurality of processor-executable instructions to:
receive a noise vector;
receive reference data;
select one or more transformations to apply to the reference data;
apply the selected one or more transformations to the reference data;
compare the reference data, with the one or more applied transformations, to the noise vector;
iteratively repeat the selecting, applying, and comparing, until the reference data, with a particular set of applied transformations, matches the noise vector within a first threshold measure of similarity;
receive test data;
apply the particular set of applied transformations to the test data; and
determine whether the test data matches the reference data based on whether the test data, with the particular set of applied transformations, matches the noise vector within a second measure of similarity.
9. The non-transitory computer-readable medium of claim 8 , wherein the determining whether the test data matches the reference data is performed without comparing the test data to the reference data.
10. The non-transitory computer-readable medium of claim 8 , wherein the test data includes biometric information received from a biometric authentication component of a User Equipment (“UE”).
11. The non-transitory computer-readable medium of claim 10 , wherein the plurality of processor-executable instructions further include processor-executable instructions to:
determine whether to authenticate a user, associated with the biometric information, based on whether the test data matches the reference data.
12. The non-transitory computer-readable medium of claim 10 , wherein the biometric information includes at least one of:
image data,
fingerprint data, or
audio data.
13. The non-transitory computer-readable medium of claim 8 , wherein the iteratively repeated selecting and applying are performed by a generator component of a Generative Adversarial Network (“GAN”), and wherein the iteratively repeated comparing is performed by a discriminator component of the GAN.
14. The non-transitory computer-readable medium of claim 8 , wherein the noise vector is different from one or more other noise vectors used to determine whether test data matches other reference data.
15. A method, comprising:
receiving a noise vector;
receiving reference data;
selecting one or more transformations to apply to the reference data;
applying the selected one or more transformations to the reference data;
comparing the reference data, with the one or more applied transformations, to the noise vector;
iteratively repeating the selecting, applying, and comparing, until the reference data, with a particular set of applied transformations, matches the noise vector within a first threshold measure of similarity;
receiving test data;
applying the particular set of applied transformations to the test data; and
determining whether the test data matches the reference data based on whether the test data, with the particular set of applied transformations, matches the noise vector within a second measure of similarity.
16. The method of claim 15 , wherein the determining whether the test data matches the reference data is performed without comparing the test data to the reference data.
17. The method of claim 15 , wherein the test data includes biometric information received from a biometric authentication component of a User Equipment (“UE”).
18. The method of claim 17 , the method further comprising:
determining whether to authenticate a user, associated with the biometric information, based on whether the test data matches the reference data.
19. The method of claim 15 , wherein the iteratively repeated selecting and applying are performed by a generator component of a Generative Adversarial Network (“GAN”), and wherein the iteratively repeated comparing is performed by a discriminator component of the GAN.
20. The method of claim 15 , wherein the noise vector is different from one or more other noise vectors used to determine whether test data matches other reference data.