Spoof detection using head pose to eye gaze correspondence
Methods, systems, and computer-readable storage media for determining that a subject is a live person include providing instructions for performing a head movement, capturing a set of images of a subject subsequent to providing the instructions to perform the head movement, determining, from the set of images, a first metric indicative of an amount of change in eye gaze directions of the subject during the head movement, determining, from the set of images, a second metric indicative of a change in head positions of the subject during the head movement, determining, based at least on the first metric and the second metric, a third metric indicative of a likelihood that the subject is a live person, determining that the third metric satisfies a first threshold condition, and in response to determining that the third metric satisfies the first threshold condition, identifying the subject as a live person.
1. A computer-implemented method comprising:
providing, by a processor of a user device having a camera, instructions on a display panel of the user device, the instructions provided for performing a head movement;
capturing a set of images of a subject using the camera of the user device subsequent to providing the instructions to perform the head movement;
determining, by the processor from the set of images, a first metric indicative of an amount of change in eye gaze directions of the subject during the head movement;
determining, by the processor from the set of images, a second metric indicative of a change in head positions of the subject during the head movement;
determining, by the processor based at least on the first metric and the second metric, a third metric indicative of a likelihood that the subject is a live person, wherein the third metric is a weighted sum of (i) the second metric and (ii) an inverse of the first metric;
determining that the third metric satisfies a first threshold condition; and
in response to determining that the third metric satisfies the first threshold condition, identifying the subject as a live person.
2. The computer-implemented method of claim 1 , wherein the instructions provide an optical target that is displayed by a device screen for the subject to look at, while the set of images is being captured.
3. The computer-implemented method of claim 1 , wherein the first metric is substantially proportional to the amount of change in eye gaze directions.
4. The computer-implemented method of claim 1 , wherein the second metric is substantially proportional to an amount of change in the head positions during the head movement.
5. The computer-implemented method of claim 1 , wherein the third metric is a weighted sum of (i) the second metric and (ii) the first metric.
6. The computer-implemented method of claim 1 , wherein the instructions to perform the head movement comprise instructions to perform one of: a yaw, a pitch or a roll of a head of the subject.
7. The computer-implemented method of claim 1 , further comprising:
determining a presence of at least one transitional state as a head position changes from an initial head position to a final head position; and
determining that the head movement between the initial head position and the final head position exceeds a second threshold condition.
8. A non-transitory, computer-readable medium storing one or more instructions executable by a computer system to perform operations comprising:
providing, by a processor of a user device having a camera, instructions on a display panel of the user device, the instructions provided for performing a head movement;
capturing a set of images of a subject using the camera of the user device subsequent to providing the instructions to perform the head movement;
determining, by the processor from the set of images, a first metric indicative of an amount of change in eye gaze directions of the subject during the head movement;
determining, by the processor from the set of images, a second metric indicative of a change in head positions of the subject during the head movement;
determining, by the processor based at least on the first metric and the second metric, a third metric indicative of a likelihood that the subject is a live person, wherein the third metric is a weighted sum of (i) the second metric and (ii) an inverse of the first metric;
determining that the third metric satisfies a first threshold condition; and
in response to determining that the third metric satisfies the first threshold condition, identifying the subject as a live person.
9. The non-transitory, computer-readable medium of claim 8 , wherein the instructions provide an optical target that is displayed by a device screen for the subject to look at, while the set of images is being captured.
10. The non-transitory, computer-readable medium of claim 8 , wherein the first metric is substantially proportional to the amount of change in eye gaze directions.
11. The non-transitory, computer-readable medium of claim 8 , wherein the second metric is substantially proportional to an amount of change in the head positions during the head movement.
12. The non-transitory, computer-readable medium of claim 8 , wherein the third metric is a weighted sum of (i) the second metric and (ii) the first metric.
13. A computer-implemented system comprising:
one or more computers; and
one or more computer memory devices interoperably coupled with the one or more computers and having tangible, non-transitory, machine-readable media storing one or more instructions that, when executed by the one or more computers, perform operations comprising:
providing, by a processor of a user device having a camera, instructions on a display panel of the user device, the instructions provided for performing a head movement;
capturing a set of images of a subject using the camera of the user device subsequent to providing the instructions to perform the head movement;
determining, by the processor from the set of images, a first metric indicative of an amount of change in eye gaze directions of the subject during the head movement;
determining, by the processor from the set of images, a second metric indicative of a change in head positions of the subject during the head movement;
determining, by the processor based at least on the first metric and the second metric, a third metric indicative of a likelihood that the subject is a live person, wherein the third metric is a weighted sum of (i) the second metric and (ii) an inverse of the first metric;
determining that the third metric satisfies a first threshold condition; and
in response to determining that the third metric satisfies the first threshold condition, identifying the subject as a live person.
14. The computer-implemented system of claim 13 , wherein the instructions provide an optical target that is displayed by a device screen for the subject to look at, while the set of images is being captured.
15. The computer-implemented system of claim 13 , wherein the first metric is substantially proportional to the amount of change in eye gaze directions.
16. The computer-implemented system of claim 13 , wherein the second metric is substantially proportional to an amount of change in the head positions during the head movement.
17. The computer-implemented system of claim 13 , wherein the third metric is a weighted sum of (i) the second metric and (ii) the first metric.