| 摘要 |
In general, liveness detection techniques are described for facial recognition. The techniques enable potential detection and mitigation of attempts to authenticate by spoofing. An example method includes determining, by a computing device, a challenge pattern against which to match an authentication input to detect liveness, and displaying, using a display device, a graphical user interface (GUI) including an element and moving the element according to the challenge pattern within the GUI. The method further includes receiving, from an image capture device, at least a first image of a face and a second image of the face, and detecting one or more eye movements based on the first and second images of the face. The method further includes determining whether to deny authentication with respect to accessing one or more functionalities controlled by the computing device. |
| 主权项 |
1. A method comprising:
determining, by a computing device, a challenge pattern against which to match an authentication input to detect liveness; displaying, using a display device coupled to the computing device, a graphical user interface including an element; moving the element according to the challenge pattern within the graphical user interface; receiving, from an image capture device coupled to the computing device, at least a first image of a face and a second image of the face; detecting, by the computing device, one or more eye movements based at least in part on the first image of the face and the second image of the face, at least in part by:
detecting, by the computing device, a first set of gradients among a first plurality of subimages extracted from the first image and a second set of gradients among a second plurality of subimages extracted from the second image, each subimage of the first plurality of subimages and the second plurality of subimages including a respective plurality of pixels;generating, by the computing device, a respective histogram of gradients corresponding to each respective subimage of the first plurality of subimages extracted from the first image and the second plurality of subimages extracted from the second image;determining, by the computing device, an angle associated with each respective pixel of each respective subimage of the first plurality of subimages and the second plurality of subimages, each determined angle representing a color transition between the respective pixel and a respective adjacent pixel;based on the determined angle associated with each respective pixel, assigning, by the computing device, a single angle category selected from a plurality of angle categories to each respective pixel; andgenerating, by the computing device, a plurality of votes associated with each respective histogram of gradients, each vote of the plurality of votes based at least in part on the respective angle category of the corresponding histogram of gradients; and determining, by the computing device, whether to deny authentication with respect to accessing one or more functionalities controlled by the computing device, based at least in part on the detected one or more eye movements. |
