- Patent Number:
11710,297
- Appl. No:
16/752634
- Application Filed:
January 25, 2020
- Abstract:
A system and method for detecting a potential match between a candidate facial image and a dataset of facial images is described. Some implementations of the invention determine whether a candidate facial image (or multiple facial images) of a person taken, for example, at point of entry corresponds to one or more facial images stored in a dataset of persons of interest (e.g., suspects, criminals, terrorists, employees, VIPs, “whales,” etc.). Some implementations of the invention detect potential fraud in a dataset of facial images. In a first form of potential fraud, a same facial image is associated with multiple identities. In a second form of potential fraud, different facial images are associated with a single identity, as in the case, for example, of identity theft. According to various implementations of the invention, spectral clustering techniques are used to determine a likelihood that pairs of facial images (or pairs of facial image sets) correspond to the person or different persons.
- Inventors:
Aeva, Inc. (Mountain View, CA, US)
- Assignees:
Aeva, Inc. (Mountain View, CA, US)
- Claim:
1. A method for detecting a potential match between a probe and a plurality of entries in a dataset, wherein each entry in the dataset comprises a plurality of gallery images, the method comprising: receiving the probe by a computing platform, the probe comprising a plurality of probe images; for each respective entry in the dataset: spectrally clustering, via the computing platform, the plurality of probe images and the plurality of gallery images of the respective entry to determine whether the plurality of probe images and the plurality of gallery images collectively correspond to one cluster or two clusters by evaluating a hypothesis test with only two hypotheses including a first hypothesis that the plurality of probe images and the plurality of gallery images collectively correspond to one cluster, and a second hypothesis that the plurality of probe images and the plurality of gallery images collectively correspond to two clusters, when the plurality of probe images and the plurality of gallery images collectively correspond to one cluster, identifying the probe and the respective entry as a match in the dataset, and when the plurality of probe images and the plurality of gallery images collectively correspond to two clusters, identifying the probe as unique in the dataset.
- Claim:
2. The method of claim 1 , wherein spectrally clustering the plurality of probe images and the plurality of gallery images comprises: forming an adjacency matrix of biometric scores of a size (N1+N2) by (N1+N2), wherein N1 is a number of probe images and wherein N2 is a number of gallery images, determining a graph Laplacian based on the adjacency matrix, determining an eigenspace decomposition, including eigenvalues and eigenvectors, based on the graph Laplacian, and estimating a number of clusters based on the eigenspace.
- Claim:
3. The method of claim 1 , wherein spectrally clustering the plurality of probe images and the plurality of gallery images comprises: assigning each of the plurality of probe images to an individual vertex in a graph; assigning each of the plurality of gallery images to an individual vertex in the graph; and determining a similarity score for each pair of vertices in the graph.
- Claim:
4. The method of claim 2 , wherein determining a graph Laplacian comprises: determining the graph Laplacian as L=D−W.
- Claim:
5. The method of claim 2 , wherein determining a graph Laplacian comprises: determining the graph Laplacian as L=I−D −1 W.
- Claim:
6. The method of claim 2 , wherein determining a graph Laplacian comprises: determining the graph Laplacian as L=I−D −1/2 WD −1/2 .
- Claim:
7. The method of claim 2 , wherein forming an adjacency matrix comprises: determining a similarity score between one of the plurality of probe images and one of the plurality of gallery images.
- Claim:
8. The method of claim 2 , wherein forming an adjacency matrix comprises: determining a similarity score between each pair of images in a set of images comprised of the plurality of probe images and the plurality of gallery images.
- Claim:
9. The method of claim 1 , wherein the hypothesis test is expressed as [mathematical expression included] wherein f(Λ, V) is a general hypothesis function of a graph Laplacian formed from the plurality of probe images and the plurality gallery image, the graph Laplacian having eigenvalues, Λ, and eigenvectors, V; wherein H 0 is the second hypothesis that the plurality of probe images and the plurality of gallery images collectively correspond to two clusters; wherein H 1 is the first hypothesis that the plurality of probe images and the plurality of gallery images collectively corresponding to one cluster; and wherein η is a threshold selected to satisfy one or more performance criteria.
- Claim:
10. The method of claim 9 , wherein the threshold is a negative number.
- Claim:
11. A method for detecting a potential match between a probe and a plurality of entries in a dataset, wherein each entry in the dataset comprises a plurality of gallery biometrics, the method comprising: receiving the probe by a computing platform, the probe comprising a plurality of probe biometrics; for each of the plurality of entries in the dataset: spectrally clustering, via the computing platform, the plurality of probe biometrics and the plurality of gallery biometrics of the entry to determine whether the plurality of probe biometrics and the plurality of gallery biometrics collectively correspond to one cluster or two clusters by evaluating a hypothesis test with only two hypotheses including a first hypothesis that the plurality of probe biometrics and the plurality of gallery biometrics collectively correspond to one cluster, and a second hypothesis that the plurality of probe biometrics and the plurality of gallery biometrics collectively correspond to two clusters, when the plurality of probe biometrics and the plurality of gallery biometrics collectively correspond to one cluster, identifying the probe and the respective entry as a match in the dataset, and when the plurality of probe biometrics and the plurality of gallery biometrics collectively correspond to two clusters, identifying the probe as unique in the dataset.
- Claim:
12. The method of claim 11 , wherein spectrally clustering the plurality of probe biometrics and the plurality of gallery biometrics comprises: forming an adjacency matrix of biometric scores of a size (N1+N2) by (N1+N2), wherein N1 is a number of probe biometrics and wherein N2 is a number of gallery biometrics, determining a graph Laplacian based on the adjacency matrix, determining an eigenspace decomposition, including eigenvalues and eigenvectors, based on the graph Laplacian, and estimating a number of clusters based on the eigenspace.
- Claim:
13. The method of claim 11 , wherein the plurality of probe biometrics comprises a first biometric type and a second biometric type, wherein the plurality of gallery biometrics comprises the first biometric type and the second biometric type, and wherein the first biometric type and the second biometric type are different from one another.
- Claim:
14. The method of claim 11 , wherein the plurality of probe biometrics comprises biometric representations of a processed image, a fingerprint, a palmprint, an iris scan, a 3D mesh, a genetic sequence, a heartbeat, a gait or a speech component.
- Claim:
15. The method of claim 11 , wherein the plurality of probe biometrics is divided into separate homogeneous biometrics, the spectral clustering is performed for each biometric, and the results are combined, to improve performance.
- Claim:
16. The method of claim 15 , wherein the combination is done in the eigenspace for each biometric or related component.
- Claim:
17. The method of claim 15 , wherein the combination is done with a combination of the separate adjacency matrices for each biometric or related component.
- Claim:
18. The method of claim 15 , wherein the combination is done on the resulting clusters, or a function of the clusters, for each biometric or related component.
- Claim:
19. The method of claim 11 , wherein the hypothesis test is expressed as [mathematical expression included] wherein f(Λ, V) is a general hypothesis function of a graph Laplacian formed from the plurality of probe biometrics and the plurality gallery biometrics, the graph Laplacian having eigenvalues, Λ, and eigenvectors, V; wherein H 0 is the second hypothesis that the plurality of probe biometrics and the plurality of gallery biometrics collectively correspond to two clusters; wherein H 1 is the first hypothesis that the plurality of probe biometrics and the plurality of gallery biometrics collectively corresponding to one cluster; and wherein η is a threshold selected to satisfy one or more performance criteria.
- Claim:
20. The method of claim 19 , wherein the threshold is a negative number.
- Patent References Cited:
8085993 December 2011 Jacobson
20030041041 February 2003 Cristianini
20040013286 January 2004 Viola
20040258284 December 2004 Daniel
20060063156 March 2006 Willman
20070253624 November 2007 Becker
20080275862 November 2008 Liu
20090028433 January 2009 Tolliver
20090253102 October 2009 Porikli
20100208951 August 2010 Williams
20100329546 December 2010 Smith
20130010154 January 2013 Maeda
20130266193 October 2013 Tiwari
20140044322 February 2014 Kumar
20140072208 March 2014 Kenyon
20140330819 November 2014 Raina
20150186752 July 2015 Xu
- Primary Examiner:
Wong, Leslie
- Attorney, Agent or Firm:
Toering Patents PLLC
- Accession Number:
edspgr.11710297
Processing Request
No Comments.