Verification method and system
The present disclosure relates to a method of determining whether a live human face is present for use as part of a facial recognition system and/or method, whereby a moving stimulus is generated. Human face movement in response to the generated stimulus is predicted using a model. The stimulus is presented to a face of a person. A movement of the face in response to the stimulus is tracked using a camera, and it is determined whether a live human face is present by comparing the movement of the face against the prediction.
1. A method of determining whether a live human face is present, comprising:
generating a moving stimulus;
predicting, using a model, human face movement in response to the presentation of said generated moving stimulus on a screen, wherein the model models a predicted correspondence distance between a modelled location of a gaze of a human face on the screen and a location of the said generated moving stimulus on the screen at a particular time;
presenting the generated moving stimulus on the screen to a presented human face;
using a camera for tracking a movement of the presented human face in B response to the generated moving stimulus presented on the screen, wherein tracking the movement of the presented human face comprises:
determining a tracked location of the gaze of the presented human face on the screen;
tracking a movement of the tracked location of the gaze of the presented human face; and
calculating a correspondence distance between the tracked location of the gaze of the presented human face and the location of the generated moving stimulus on the screen at a particular time; and
determining whether the presented human face is a live human face by comparing the predicted human face movement against said tracked movement of the presented human face.
2. A method according to claim 1 , wherein the model represents live human face movement in response to the said generated moving stimulus.
3. A method according to claim 1 , wherein the generated moving stimulus is generated in dependence on a varying parameter.
4. A method according to claim 3 , further comprising generating the varying parameter in dependence on one or more dynamic properties.
5. A method according to claim 4 , wherein the one or more dynamic properties are properties of a user device adapted to implement the method; optionally wherein the one or more properties comprise one or more of: clock time; location; orientation; rotation; and acceleration.
6. A method according to claim 3 , wherein the varying parameter comprises a random element.
7. A method according to claim 1 , wherein the generated moving stimulus is a visual stimulus.
8. A method according to claim 7 , wherein the color, luminance and/or orientation of the generated moving stimulus is generated in dependence on the varying parameter.
9. A method according to claim 7 , wherein the generated moving stimulus varies in size as it moves.
10. A method according to claim 1 , wherein the generated moving stimulus moves continuously.
11. A method according to claim 1 , wherein the generated moving stimulus moves along a path, the path being generated in dependence on the varying parameter; optionally wherein a starting location and/or ending location of the path are generated in dependence on the varying parameter.
12. A method according to claim 1 , wherein the method further comprises comparing the correspondence distance against the predicted correspondence distance for a particular time.
13. A method according to claim 12 , wherein calculating the correspondence distance is performed at the same time as presenting the generated moving stimulus to the face.
14. A method according to claim 1 , wherein one or more properties of the generated moving stimulus are used as inputs for the model.
15. A method according to claim 1 , wherein the model comprises historic data; optionally related to the presented human face to which the generated moving stimulus is presented.
16. A computer program product, stored on a non-transitory computer readable medium, comprising software code adapted to carry out the method of claim 1 .
17. A system for determining whether a live human face is present, the system comprising:
a display for displaying a generated moving stimulus;
a camera for capturing visual data of a presented human face;
a data store for storing a model of live human face movement; and
a processor adapted to execute the computer program product of claim 16 .
18. A system for determining whether a live human face is present, comprising:
a user device, comprising:
a module adapted to generate a moving stimulus;
a module adapted to present the generated moving stimulus on a screen to a presented human face;
a module adapted to track a movement of the presented human face in response to the generated moving stimulus which is presented on the screen by:
determining a tracked location of a gaze of the presented human face on the screen;
tracking a movement of the location of the gaze of the presented human face; and
calculating a correspondence distance between the tracked location of the gaze of the presented human face and the location of the generated moving stimulus on the screen at a particular time; and
a remote prediction module adapted to predict, using a model, human face movement in response to the presentation of said generated moving stimulus, wherein the model models a predicted correspondence distance between a modelled location of a gaze of a human face on the screen and a location of the said generated moving stimulus on the screen at a particular time;
a remote determination module adapted to determine whether the presented human face is a live human face by comparing the tracked movement of the presented human face against said predicted human face movement.