Authentication based on makeup application technique
Techniques are disclosed for authenticating a person based on the person's individual makeup application techniques (MATs) for secured access purposes. As combinations of MATs are highly individualized and may even be unique between different people, the disclosed techniques leverage MAT signatures as an authentication factor. For example, various makeup application technique footprints produced by corresponding MATs which were utilized to apply makeup to a subject person may be extracted from a captured image in which the subject person is depicted. One or more combinations of the extracted footprints may be compared to a MAT signature registry to thereby authorize the identity of the subject person or otherwise deny authorization of the subject person. MAT-based authentication may be combined with other authentication factors (such as biometric identification techniques) to provide increased levels of security.
1 . A method of verifying an identity of a person based on a makeup application technique signature, the method comprising:
capturing, by one or more computing devices and by utilizing an image capturing device, an image of the person;
applying one or more image processing and/or computer vision techniques to the captured image, thereby distinguishing a set of makeup application techniques (MATs) via which make-up was applied to the person depicted in the captured image, the distinguished set of MATs including at least one of a directionality of application, a pressure pattern, a region of application, a color distribution, or a shading pattern;
providing a plurality of makeup application technique (MAT) signatures, each MAT signature of the plurality of MAT signatures corresponding to a respective person and indicative of a respective combination of techniques via which the respective person applies makeup to themselves, and each MAT signature approved by the respective person;
determining, by the one or more computing devices, whether a respective distance between the distinguished set of MATs corresponding to the person depicted in the captured image and each MAT signature of the plurality of MAT signatures is less than a threshold distance;
when none of the respective distances is less than the threshold distance, indicating, by the one or more computing devices and via at least one of a display or a communication interface, that a match was not found between the distinguished set of MATs corresponding to the person depicted in the captured image and any of the plurality of MAT signatures; and
when one or more of the respective distances is less than the threshold distance:
selecting, by the one or more computing devices, a MAT signature corresponding to a least distance among the one or more respective distances;
indicating, via at least one of the display or the communication interface, a match between the distinguished set of MATs corresponding to the person depicted in the captured image and the selected MAT signature, the indicating of the match utilized as a first authentication factor; and
utilizing another authentication factor in combination with the first authentication factor to authenticate the person depicted in the captured image, the another authentication factor including at least one of a password, a personal identification number, facial recognition, another biometric characteristic, or a prompt sent to another device.
2 . The method of claim 1 , wherein the respective combination of techniques via which the respective person applies makeup includes at least one of:
a utilization of a particular set of makeup tools;
an order of application of a particular makeup product to multiple areas of the person;
an order of application of different makeup products;
respective directionalities of application of one or more makeup products; or
respective pressures of application of the one or more makeup products.
3 . The method of claim 2 , wherein the respective combination of techniques via which the respective person applies makeup further includes at least one of:
a type of makeup product;
a formulation of the type of makeup product;
a combination of types of makeup products;
a combination of types and formulations of more than one makeup product;
respective amounts utilized of the one or more makeup products;
respective shapes of application of the one or more makeup products;
a color of a specific makeup product;
a combination of colors of two or more specific makeup products; or
one or more regions of a face of the respective person to which makeup is applied.
4 . The method of claim 1 , wherein the one or more image processing and/or computer vision techniques include at least one of: edge detecting, color segmenting, or utilizing another type of feature extraction algorithm.
5 . The method of claim 1 , wherein capturing the image of the person includes capturing an image of a portion of the person other than a face of the person and to which makeup has been applied.
6 . The method of claim 1 , wherein the another authentication factor is the facial recognition or the another biometric characteristic.
7 . The method of claim 1 , wherein:
a first instance of the method is executed during a session and the person depicted in the captured image is identified as being the respective person corresponding to the selected MAT signature based on the execution of the first instance; and
the method further comprises subsequently executing additional instances of the method during the session to thereby providing continual verification of the identity of the person during the session.
8 . The method of claim 1 , further comprising updating the selected MAT signature based on the distinguished set of MATs.
9 . The method of claim 1 , further comprising modifying or deleting, based on an instruction provided by the person, one or more MAT signatures that are included in the plurality of MAT signatures and that correspond to the person.
10 . The method of claim 1 , wherein the plurality of MAT signatures includes one or more MAT signatures of the person, and the method further comprises generating at least one of the one or more MAT signatures of the person by image processing multiple images of the person.
11 . The method of claim 10 , wherein generating the at least one of the one or more MAT signatures of the person includes discovering the at least one of the one or more MAT signatures of the person by applying one or more unsupervised discovery techniques to data generated from the image processing of the multiple images of the person.
12 . A system for verifying an identity of a person based on a makeup application technique signature, the system comprising:
one or more processors;
a plurality of makeup application technique (MAT) signatures stored in one or more data stores of the system, each makeup application technique signature of the plurality makeup application technique signatures corresponding to a respective person and indicative of a respective combination of techniques via which the respective person applies makeup to themselves, wherein each MAT signature is approved by the respective person; and
one or more non-transitory computer-readable media coupled to the one or more processors and storing instructions thereon that, when executed by the one or more processors, cause the system to:
capture, via an image capturing device, an image of the person;
apply one or more image processing and/or computer vision techniques to the captured image to thereby distinguish a set of makeup application techniques (MATs) via which make-up was applied to the person depicted in the captured image, the distinguished set of MATs including at least one of a directionality of application, a pressure pattern, a region of application, a color distribution, or a shading pattern;
determine whether a respective distance between the distinguished set of MATs corresponding to the person depicted in the captured image and each MAT signature of the plurality of MAT signatures is less than a threshold distance;
when none of the respective distances is less than the threshold distance, indicate, via least one of a display or a communication interface, that a match was not found between the distinguished set of MATs corresponding to the person depicted in the captured image and any of the plurality of MAT signatures; and
when one or more of the respective distances is less than the threshold distance:
select a MAT signature corresponding to a least distance among the one or more respective distances:
indicate, via at least one of the display or the communication interface, a match between the distinguished set of MATs corresponding to the person depicted in the captured image and the selected MAT signature, the indication of the match utilized as a first authentication factor; and
utilize another authentication factor in combination with the first authentication factor to authenticate the person depicted in the captured image, the another authentication factor including at least one of a password, a personal identification number, facial recognition, another biometric characteristic, or a prompt sent to another device.
13 . The system of claim 12 , wherein the respective combination of techniques via which the respective person applies makeup includes at least one of:
a utilization of a particular set of makeup tools;
an order of application of a particular makeup product to multiple areas of the person;
an order of application of different makeup products;
respective directionalities of application of one or more makeup products; or
respective pressures of application of the one or more makeup products.
14 . The system of claim 13 , wherein the respective combination of techniques via which the respective person applies makeup further includes at least one of:
a type of makeup product;
a formulation of the type of makeup product;
a combination of types of makeup products;
a combination of types and formulations of more than one makeup product;
respective amounts utilized of the one or more makeup products;
respective shapes of application of the one or more makeup products;
a color of a specific makeup product;
a combination of colors of two or more specific makeup products; or
one or more regions of a face of the respective person to which makeup is applied.
15 . The system of claim 12 , wherein the plurality of MAT signatures includes more than one signature corresponding to the person.
16 . The system of claim 12 , wherein the one or more image processing and/or computer vision techniques include at least one of: edge detection, color segmentation, or another type of feature extraction algorithm.
17 . The system of claim 12 , wherein the plurality of MAT signatures includes one or more MAT signatures of the person, and the instructions, when executed by the one or more processors, cause the system further to generate at least one of the one or more MAT signatures of the person by image processing multiple images of the person.
18 . The system of claim 17 , wherein the generation of the at least one of the one or more MAT signatures of the person includes a discovery, via one or more unsupervised discovery techniques applied to data generated by the image processing of the multiple images of the person, of the at least one of the one or more MAT signatures of the person.