Systems and methods for biometric identity and authentication

Patent metadata

Field	Value
Publication number	US-11449587-B2
Application number	US-201816194867-A
Country	US
Kind code	B2
Filing date	Nov 19, 2018
Priority date	Nov 20, 2017
Publication date	Sep 20, 2022
Grant date	Sep 20, 2022

Abstract

In accordance with some embodiments, an apparatus for privacy protection is provided. The apparatus includes a housing arranged to hold a second device; one or more sensors, at least partially supported by the housing, operable to continuously collect biometric data of a user; an authentication neural network, operable to extract from the biometric data a plurality feature vectors associated with a plurality of identifiable scores; and a decision unit, coupled to the authentication neural network, operable to generate an authentication score as a function of the plurality of identifiability scores and the plurality of feature vectors, determine whether or not the authentication score satisfies an authentication threshold, and gate electronic access to the second device base on whether or not the authentication score satisfies the authentication threshold.

First claim

What is claimed is:

1. An apparatus comprising:
a housing arranged to hold a second device, distinct from the apparatus;
one or more sensors, at least partially supported by the housing, operable to obtain at least a portion of biometric data of a user of the second device;
a communication interface connectable to the second device;
a communication device at least partially supported by the housing, operable to receive parameters of an authentication neural network from a server through a secure channel, wherein the parameters are generated by the server upon training a remote neural network using biometric data associated with a group of different users;
the authentication neural network, operable to extract from the biometric data a plurality of feature vectors associated with a plurality of identifiability scores, wherein each of the plurality of identifiability scores provides a quantitative characterization of a relative uniqueness of a corresponding one of the plurality of feature vectors among the group of different users, wherein the parameters of the authentication neural network are generated by the server based on training vectors unique within the group from a group of apparatuses each arranged to hold a respective user equipment used by a respective user in the group of different users; and
a decision unit, coupled to the authentication neural network, operable to generate an authentication score as a function of the plurality of identifiability scores and the plurality of feature vectors, determine whether or not the authentication score satisfies an authentication threshold, and gate electronic access of the user to the second device via the communication interface based on whether or not the authentication score satisfies the authentication threshold.

2. The apparatus of claim 1 further comprising a plurality of conditioning units, coupled to the one or more sensors, operable to condition the biometric data to remove outliers and normalize the biometric data.

3. The apparatus of claim 2 , wherein the authentication neural network includes an input layer to receive the biometric data from the plurality of conditioning units.

4. The apparatus of claim 1 , wherein the authentication neural network includes a layer to extract the plurality of feature vectors based on the parameters.

5. The apparatus of claim 1 , wherein the second device includes one or more device sensors to obtain a set of biometric data to be included in the biometric data, and communicates the set of biometric data through the communication interface of the apparatus.

6. The apparatus of claim 1 , wherein the the apparatus is further operable to, via one or more of the communication interface and the communication device:
send the biometric data to the remote neural network through the secure channel;
receive feature vectors extracted by the remote neural network; and
forward the feature vectors extracted by the remote neural network to the decision unit for comparison.

7. The apparatus of claim 6 , wherein the decision unit includes:
a comparator to compare the feature vectors with the plurality of feature vectors to derive a difference between the feature vectors and the plurality of feature vectors; and
an evaluator to indicate whether the authentication score is stable based on the difference.

8. The apparatus of claim 1 , wherein the second device includes:
a display to present candidate features to a user for selection, wherein sensors on at least one of the second device or the apparatus are capable of detecting the candidate features; and
one or more input devices to obtain a set of user-selected features from the candidate features, wherein the set of user-selected features is communicated to the apparatus via the communication interface of the apparatus.

9. The apparatus of claim 1 , wherein the authentication neural network receives updates of the parameters from the server through the communication device of the apparatus.

10. The apparatus of claim 1 further comprising a key generator to derive an authentication key from the plurality of feature vectors and provide the authentication key to the decision unit for generating the authentication score as the function of the plurality of identifiability scores and the authentication key.

11. The apparatus of claim 1 further comprising a secure storage to store an identifier of the apparatus, wherein the decision unit retrieves the identifier from the secure storage and generates the authentication score as the function of the plurality of identifiability scores, the plurality of feature vectors, and the identifier.

12. A method comprising:
at a controller of an apparatus holding a second device distinct from the apparatus, wherein the apparatus includes a housing arranged to hold the second device, one or more sensors at least partially supported by the housing, a communication interface connectable to the second device, and a communication device:
obtaining, via the one or more sensors, at least a portion of biometric data of a user of the second device;
receiving, by the communication device at least partially supported by the housing, parameters of an authentication neural network from a server through a secure channel, wherein the parameters are generated by the server upon training a remote neural network using biometric data associated with a group of different users;
extracting, by the authentication neural network, from the biometric data a plurality of feature vectors associated with a plurality of identifiability scores, wherein each of the plurality of identifiability scores provides a quantitative characterization of a relative uniqueness of a corresponding one of the plurality of feature vectors among the group of different users, wherein the parameters of the authentication neural network are generated by the server based on training vectors unique within the group from a group of apparatuses each arranged to hold a respective user equipment used by a respective user in the group of different users;
generating an authentication score as a function of the plurality of identifiability scores and the plurality of feature vectors;
determining whether or not the authentication score satisfies an authentication threshold; and
gating electronic access of the user to the second device via the communication interface based on whether or not the authentication score satisfies the authentication threshold.

13. The method of claim 12 , further comprising extracting the plurality of feature vectors based on the parameters.

14. The method of claim 12 , further comprising receiving, via the communication interface, a set of biometric data to be included in the biometric data from the second device.

15. The method of claim 12 further comprising:
sending the biometric data to the remote neural network through the secure channel; and
receiving feature vectors extracted by the remote neural network.

16. The method of claim 15 , further comprising:
comparing the feature vectors with the plurality of feature vectors to derive a difference between the feature vectors and the plurality of feature vectors; and
providing an indication of whether the authentication score is stable based on the difference.

17. The method of claim 12 , further comprising:
causing the second device to present candidate features to the user for selection, wherein sensors on at least one of the second device or the apparatus are capable of detecting the candidate features; and
receiving from the second device a set of user-selected featur

Assignees

• Ppip Llc

Inventors

• Fong Michael
• Fong Neric Hsin-Wu
• Chan Stephen Kuo-Tung Seay
• Tokola Krishna Teja
• Le Huyen Nguyen Ngoc Cam

Classifications

• G06V40/15

  Biometric patterns based on physiological signals, e.g. heartbeat, blood flow · CPC title

• G06V10/82

  using neural networks · CPC title

• G06V10/454

  Integrating the filters into a hierarchical structure, e.g. convolutional neural networks [CNN] · CPC title

• G06F18/217

  Validation; Performance evaluation; Active pattern learning techniques · CPC title

• G06F18/256

  of results relating to different input data, e.g. multimodal recognition · CPC title