Liveness Detection
AssigneeAccenture Global Services Limited
InventorAlessio Cavallini
Life of this patent
Estimating the term…
Abstract
The present disclosure concerns a method of verifying the presence of a living face in front of a camera (112), the method including: capturing by said camera a sequence of images of a face; detecting a plurality of features of said face in each of said images; measuring parameters associated with said detected features to determine whether each of a plurality of liveness indicators is present in said images; determining whether or not said face is a living face based on the presence in said images of a combination of at least two of said liveness indicators.
Claims21
(canceled)
A computer-implemented method comprising:
- obtaining training data for training a machine to output a threshold value for each of multiple facial features, the training data including (i) multiple facial images and, (ii) for each facial image, an indication of whether the facial image is of a picture of a person in front of a camera or of a person actually positioned in front of a camera;
- training the machine using the (i) multiple facial images and, (ii) for each facial image, an indication of whether the facial image is of a picture or of a person actually positioned in front of a camera;
- determining, using the trained machine, the threshold value for each of the multiple facial features;
- and
- using one or more of the threshold values for one or more of the facial features in determining whether one or more subsequently received facial images are of a picture of a person in front of a camera or of a person actually positioned in front of a camera.
A system comprising:
- one or more computers and one or more storage devices storing instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising:
- obtaining training data for training a machine to output a threshold value for each of multiple facial features, the training data including (i) multiple facial images and, (ii) for each facial image, an indication of whether the facial image is of a picture of a person in front of a camera or of a person actually positioned in front of a camera;
- training the machine using the (i) multiple facial images and, (ii) for each facial image, an indication of whether the facial image is of a picture or of a person actually positioned in front of a camera;
- determining, using the trained machine, the threshold value for each of the multiple facial features;
- and
- using one or more of the threshold values for one or more of the facial features in determining whether one or more subsequently received facial images are of a picture of a person in front of a camera or of a person actually positioned in front of a camera.
A non-transitory computer-readable medium storing software comprising instructions executable by one or more computers which, upon such execution, cause the one or more computers to perform operations comprising:
- obtaining training data for training a machine to output a threshold value for each of multiple facial features, the training data including (i) multiple facial images and, (ii) for each facial image, an indication of whether the facial image is of a picture of a person in front of a camera or of a person actually positioned in front of a camera;
- training the machine using the (i) multiple facial images and, (ii) for each facial image, an indication of whether the facial image is of a picture or of a person actually positioned in front of a camera;
- determining, using the trained machine, the threshold value for each of the multiple facial features;
- and
- sing one or more of the threshold values for one or more of the facial features in determining whether one or more subsequently received facial images are of a picture of a person in front of a camera or of a person actually positioned in front of a camera.
Dependencies are inferred from the claim text.