Systems and methods for privacy-enabled biometric processing
A set of distance measurable encrypted feature vectors can be derived from any biometric data and/or physical or logical user behavioral data, and then using an associated deep neural network (“DNN”) on the output (i.e., biometric feature vector and/or behavioral feature vectors, etc.) an authentication system can determine matches or execute searches on encrypted data. Behavioral or biometric encrypted feature vectors can be stored and/or used in conjunction with respective classifications, or in subsequent comparisons without fear of compromising the original data. In various embodiments, the original behavioral and/or biometric data is discarded responsive to generating the encrypted vectors. In another embodiment, distance measurable or homomorphic encryption enables computations and comparisons on cypher-text without decryption of the encrypted feature vectors. Security of such privacy enabled embeddings can be increased by implementing an assurance factor (e.g., liveness) to establish a submitted credential has not been spoofed or faked.
1. An authentication system for privacy-enabled authentication, the system comprising:
at least one processor operatively connected to a memory;
a classification component executed by the at least one processor, comprising at least a first deep neural network (“DNN”), the first DNN configured to:
accept encrypted authentication credentials, generated from a first neural network as input to the first DNN;
classify the encrypted authentication credentials during training, based on processing the encrypted authentication credentials and associated label inputs during the training;
output an array of values, from the first DNN, reflecting a probability of a first match to at least one label of the associated label inputs for identification responsive to analyzing at least one of the encrypted authentication credentials input during prediction; and
wherein the classification component is further configured to:
retrieve at least one of the encrypted authentication credentials classified during training from the memory based on identification in the array of values establishing the first match to an identity;
determine a distance between the at least one of the encrypted authentication credentials input during prediction and the at least one encrypted authentication credential retrieved; and
return a distance match, responsive to determining the distance between the encrypted authentication credential input and the at least one encrypted authentication credential meets a threshold for validating a second match to the identity.
2. The system of claim 1 , wherein the classification component is further configured to retrieve a group of authentication credentials classified during training based on identification of a group of labels having respective highest values for probability of match.
3. The system of claim 2 , wherein the classification component is further configured to determine an unknown result responsive to the distance determination not meeting the threshold for validating the second match to the identity, and the probability of the match to the at least one label not meeting a threshold probability.
4. The system of claim 1 , wherein the at least one processor is configured to build an authentication database including encrypted authentication credentials associated with a respective identification label.
5. The system of claim 1 , wherein the at least one processor is configured to enroll authentication credentials and respective labels for subsequent authentication by the first DNN.
6. The system of claim 5 , further comprising the first neural network configured to generate the encrypted authentication credentials responsive to unencrypted input of authentication credentials.
7. The system of claim 6 , wherein the at least one processor is configured to associate the encrypted authentication credentials to labels in an authentication database.
8. The system of claim 7 , wherein the at least one processor is configured to validate a plurality of encrypted authentication credentials prior to use in training the first DNN.
9. The system of claim 8 , wherein the at least one processor is configured to reject training instances of authentication credentials that exceed a threshold for validation based on distance evaluation.
10. The system of claim 1 , wherein the at least one processor is further configured to validate contemporaneous input of unencrypted authentication credentials.
11. The system of claim 10 , wherein the at least one processor is further configured to generate a liveness score based on a random set of candidate authentication instances.
12. A computer implemented method for privacy-enabled authentication with contemporaneous validation, the method comprising:
accepting, by at least one processor, encrypted authentication credentials, generated from a first neural network as input to the first DNN;
classifying, by the at least one processor, the encrypted authentication credentials during training of a first deep neural network (“DNN”), based on processing the encrypted authentication credentials and associated label inputs during the training with the first DNN;
outputting, an array of values reflecting a probability of a first match to at least one label of the associated label inputs for identification responsive to analyzing at least one of the encrypted authentication credentials input to the DNN during prediction;
retrieving, by the at least one processor, at least one of the encrypted authentication credentials classified during training based on identification in the array of values; and
determining, by the at least one processor, a distance between the at least one of the encrypted authentication credentials input during prediction and the at least one of the encrypted authentication credentials retrieved; and
returning, by the at least one processor, a distance match, responsive to determining the distance between the at least one of the encrypted authentication credential input and the at least one of the encrypted authentication credentials retrieved meets a threshold for validating a second match to the identity.
13. The method of claim 12 , wherein the act of retrieving includes retrieving a group of authentication credentials classified during training based on identification of a group of labels having respective highest values for probability of match in the array.
14. The method of claim 12 , further comprising determining an unknown result responsive to the distance comparison not meeting the threshold for validating the second match to the identity, and the probability of the match to the at least one label not meeting a threshold probability.
15. The method of claim 12 , further comprising building an authentication database including the encrypted authentication credentials associated with respective identification labels during enrollment.
16. The method of claim 12 , further comprising enrolling encrypted authentication credentials and respective labels for subsequent authentication by the first DNN.
17. The method of claim 16 , further comprising generating encrypted authentication credentials responsive to unencrypted input of authentication credentials to the first neural network.
18. The method of claim 17 , further comprising associating encrypted authentication credentials to respective labels in an authentication database.
19. The method of claim 12 , further comprising validating contemporaneous input of unencrypted authentication credentials.
20. The method of claim 19 , further comprising generating a liveness score based, at least in part, on validating a random set of candidate authentication instances.