Facial recognition database using face clustering
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for fast user enrollment for facial recognition using face clustering. A property monitoring system may reduce user input requirements by automatically performing face clustering, and requesting a user identify face images by cluster. Presenting clusters of faces to a user may allow the user to more quickly indicate multiple face images are of a user, and allow the property monitoring system to prioritize which clusters of faces are to be reviewed by the user. Accordingly, the property monitoring system may reduce processing and requests for user input by clustering faces.
1 . A computer-implemented method comprising:
identifying, from a set of face images of faces, clusters of face images, where the clusters of face images include a first cluster;
determining whether the first cluster includes images for only a single person;
in response to determining that the first cluster does not include images for only a single person:
detecting, in the first cluster, a first subcluster of face images that depict a first person and a second subcluster of face images that depict a second, different person; and
determining whether i) a number of face images in the first subcluster of face images that depict the first person satisfies an enrollment criteria for faces that were clustered with similar looking people and ii) additional images of the first person are not required to enroll the first person in a facial recognition database;
in response to determining that i) the number of face images in the first subcluster of face images that depict the first person does not satisfy the enrollment criteria for faces that were clustered with similar looking people and ii) additional images of the first person are required to enroll the first person in a facial recognition database;
detecting, from the set of face images of faces, another cluster of face images that includes additional images that depict the first person using data for the face images from the first subcluster that were indicated as showing the first person, the other cluster of faces that depict the first person including at least one face image that is not included in the first subcluster of face images that depict the first person; and
enrolling, in the facial recognition database, the first person using the other cluster of face images that includes additional images that depict the first person.
2 . The method of claim 1 , comprising:
in response to receiving an indication, determining that a second number of face images in the second subcluster of face images that depict the second person satisfies the enrollment criteria; and
in response to determining that the second number satisfies the enrollment criteria, enrolling, in the facial recognition database, the second person using the second subcluster of face images that depict the second person.
3 . The method of claim 1 , wherein enrolling the first person uses the other cluster of face images and the first subcluster of face images.
4 . The method of claim 1 , comprising providing the cluster of face images to a device; and
receiving an indication responsive to providing the cluster of face images.
5 . The method of claim 4 , wherein:
providing the cluster of face images comprises providing a cluster of unknown face images to the device; and
identifying the other cluster of face images comprises detecting, using clustering with the first subcluster of face images and in a plurality of images, the other cluster of face images that depict the first person.
6 . The method of claim 5 , wherein detecting the other cluster of face images uses a spectral clustering.
7 . The method of claim 1 , comprising providing data for the facial recognition database to a facial recognition system to cause the facial recognition system to recognize one or more people including the first person using the facial recognition database.
8 . The method of claim 1 , comprising:
recognizing, using data for the other cluster of face images in the facial recognition database, the first person; and
in response to recognizing the first person, determining whether to perform an action or to skip performing an action.
9 . The method of claim 8 , comprising, in response to recognizing the first person, sending an instruction to a door to cause the door to unlock.
10 . The method of claim 8 , comprising, in response to recognizing the first person, determining to skip sending, to a second device, an alert about the first person.
11 . The method of claim 1 , wherein identifying the other cluster comprises identifying the other cluster using at least one of an affinity matrix or a binary support vector machine.
12 . A system comprising one or more computers and one or more storage devices on which are stored instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising:
identifying, from a set of face multiple images of faces, clusters of face images, where the clusters of face images include a first cluster;
determining whether the first cluster includes images for only a single person;
in response to determining that the first cluster does not include images for only a single person:
detecting in the first cluster, a first subcluster of face images that depict a first person and a second subcluster of face images that depict a second, different person; and
determining whether i) a number of face images in the first subcluster of face images that depict the first person satisfies an enrollment criteria for faces that were clustered with similar looking people and ii) additional images of the first person are not required to enroll the first person in a facial recognition database;
in response to determining that i) the number of face images in the first subcluster of face images that depict the first person does not satisfy the enrollment criteria for faces that were clustered with similar looking people and ii) additional images of the first person are required to enroll the first person in a facial recognition database:
detecting, from the set of multiple face images of faces, another cluster of face images that includes additional images that depict the first person using data for the face images from the first subcluster that were indicated as showing the first person, the other cluster of faces that depict the first person including at least one face image that is not included in the first subcluster of face images that depict the first person; and
enrolling, in the facial recognition database, the first person using the other cluster of face images that includes the additional images that depict the first person.
13 . The system of claim 12 , the operations comprising:
in response to receiving an indication, determining that a second number of face images in the second subcluster of face images that depict the second person satisfies the enrollment criteria; and
in response to determining that the second number satisfies the enrollment criteria, enrolling, in the facial recognition database, the second person using the second subcluster of face images that depict the second person.
14 . The system of claim 12 , wherein enrolling the first person uses the other cluster of face images and the first subcluster of face images.
15 . The system of claim 12 , the operations comprising providing the cluster of face images to a device; and
receiving an indication responsive to providing the cluster of face images.
16 . The system of claim 15 , wherein:
providing the cluster of face images comprises providing a cluster of unknown face images to the device; and
identifying the other cluster of face images comprises detecting, using clustering with the first subcluster of face images and in a plurality of images, the other cluster of face images that depict the first person.
17 . A non-transitory computer storage medium encoded with instructions that, when executed by one or more computers, cause the one or more computers to perform operations comprising:
identifying, from a set of multiple face images of faces, clusters of face images, where the clusters of face images include a first cluster;
determining whether the first cluster includes images for only a single person;
in response to determining that the first cluster does not include images for only a single person:
detecting, in the first cluster, a first subcluster of face images that depict a first person and a second subcluster of face images that depict a second, different person; and
determining i) whether a number of face images in the first subcluster of face images that depict the first person satisfies an enrollment criteria for faces that were clustered with similar looking people and ii) additional images of the first person are not required to enroll the first person in a facial recognition database;
in response to determining that i) the number of face images in the first subcluster of face images that depict the first person does not satisfy the enrollment criteria for faces that were clustered with similar looking people and ii) additional images of the first person are required to enroll the first person in a facial recognition database:
detecting, from the set of multiple face images of faces, another cluster of face images that includes additional images that depict the first person using data for the face images from the first subcluster that were indicated as showing the first person, the other cluster of faces that depict the first person including at least one face image that is not included in the first subcluster of face images that depict the first person; and
enrolling, in the facial recognition database, the first person using the other cluster of face images that includes the additional images that depict the first person.
18 . The non-transitory computer storage medium of claim 17 , the operations comprising:
in response to receiving an indication, determining that a second number of face images in the second subcluster of face images that depict the second person satisfies the enrollment criteria; and
in response to determining that the second number satisfies the enrollment criteria, enrolling, in the facial recognition database, the second person using the second subcluster of face images that depict the second person.
19 . The non-transitory computer storage medium of claim 17 , wherein enrolling the first person uses the other cluster of face images and the first subcluster of face images.
20 . The non-transitory computer storage medium of claim 17 , the operations comprising providing the cluster of face images to a device; and
receiving an indication responsive to providing the cluster of face images.