Differentiating a rendered conference participant from a genuine conference participant
A videoconferencing device at an endpoint determines whether a person is a real person standing in front of a display device or if the person is instead an image rendered by a display device. In the first instance the real person will be included in a video feed for transmission to a remote endpoint. In the second instance, images of the display device on which the person is rendered will not be included in the video feed.
1. A method of differentiating a rendering of a teleconference participant from a genuine teleconference participant, comprising:
capturing, using a first optical device coupled to a processor, first image data corresponding to a first view;
establishing, using the processor, a first facial region within the first view;
determining, using the processor, a first cluster of facial feature points corresponding to the first facial region;
capturing, using a second optical device coupled to the processor, second image data corresponding to a second view;
establishing, using the processor, a second facial region within the second view;
determining, using the processor, a second cluster of facial feature points corresponding to the second facial region;
comparing, using the processor, the first cluster of facial feature points with the second cluster of facial feature points;
determining that the first cluster of facial feature points and the second cluster of facial feature points correspond to a rendered participant when a degree of similarity between the first cluster of facial feature points and the second cluster of facial feature points satisfies a first threshold; and
determining that the first cluster of facial feature points and the second cluster of facial feature points correspond to a genuine participant when the degree of similarity between the first cluster of facial feature points and the second cluster of facial feature points is below a second threshold.
2. The method of claim 1 , wherein the second threshold is different from the first threshold.
3. The method of claim 1 , further comprising transmitting, through a network interface coupled to the processor, third image data corresponding to the genuine participant, responsive to determining that the first cluster of facial feature points and the second cluster of facial feature points correspond to the genuine participant.
4. The method of claim 1 , further comprising:
establishing, using the processor, a four-sided polygon within the first view; and
determining, using the processor, that the first facial region is at least partially bounded by the four-sided polygon,
wherein determining the first cluster of facial feature points corresponding to the first facial region is responsive to determining that the first facial region is at least partially bounded by the four-sided polygon.
5. The method of claim 1 , wherein:
the first cluster of facial feature points comprises a first left-corner-of-mouth point and a first right-corner-of-mouth point; and
the second cluster of facial feature points comprises a second left-corner-of-mouth point and a second right-corner-of-mouth point.
6. The method of claim 5 , wherein:
the first cluster of facial feature points further comprises a first left-eye-point, a first right-eye-point, and a first nose-point; and
the second cluster of facial feature points further comprises a second left-eye-point, a second right-eye-point, and a second nose-point.
7. The method of claim 1 , wherein comparing, using the processor, the first cluster of facial feature points with the second cluster of facial feature points comprises determining distances between two or more facial feature points of the first cluster of facial feature points and determining distances between two or more facial feature points within the second cluster of facial feature points.
8. The method of claim 1 , wherein the first threshold is an eighty-five percent degree of similarity.
9. A videoconferencing endpoint, comprising:
a first optical device and a second optical device;
a processor coupled to the first optical device, and the second optical device; and
a memory storing instructions executable by the processor, wherein the instructions comprise instructions to:
capture, using the first optical device, first image data corresponding to a first view;
establish a first facial region within the first view;
determine a first cluster of facial feature points corresponding to the first facial region;
capture, using the second optical device, second image data corresponding to a second view;
establish a second facial region within the second view;
determine a second cluster of facial feature points corresponding to the second facial region;
compare the first cluster of facial feature points with the second cluster of facial feature points;
determine that the first cluster of facial feature points and the second cluster of facial feature points correspond to a rendered participant when a degree of similarity between the first cluster of facial feature points and the second cluster of facial feature points satisfies a first threshold; and
determine that the first cluster of facial feature points and the second cluster of facial feature points correspond to a genuine participant when the degree of similarity between the first cluster of facial feature points and the second cluster of facial feature points is below a second threshold.
10. The videoconferencing endpoint of claim 9 , wherein the second threshold is the same as the first threshold.
11. The videoconferencing endpoint of claim 9 , wherein the instructions further comprise instructions to transmit, through a network interface coupled to the processor, third image data corresponding to the genuine participant, responsive to determining that the first cluster of facial feature points and the second cluster of facial feature points correspond to the genuine participant.
12. The videoconferencing endpoint of claim 9 , wherein the instructions further comprise instructions to:
establish a four-sided polygon within the first view; and
determine that the first facial region is at least partially bounded by the four-sided polygon, and
wherein the instructions to determine the first cluster of facial feature points corresponding to the first facial region comprise instructions to determine the first cluster of facial feature points responsive to determining that the first facial region is at least partially bounded by the four-sided polygon.
13. The videoconferencing endpoint of claim 9 , wherein:
the first cluster of facial feature points comprises a first left-corner-of-mouth point and a first right-corner-of-mouth point; and
the second cluster of facial feature points comprises a second left-corner-of-mouth point and a second right-corner-of-mouth point.
14. The videoconferencing endpoint of claim 13 , wherein:
the first cluster of facial feature points further comprises a first left-eye-point, a first right-eye-point, and a first nose-point; and
the second cluster of facial feature points further comprises a second left-eye-point, a second right-eye-point, and a second nose-point.
15. The videoconferencing endpoint of claim 9 , wherein the instructions to compare the first cluster of facial feature points with the second cluster of facial feature points comprise instructions to determine distances between two or more facial feature points of the first cluster of facial feature points and determine distances between two or more facial feature points within the second cluster of facial feature points.
16. The videoconferencing endpoint of claim 9 , wherein the first threshold is an eighty-five percent degree of similarity.
17. A non-transitory computer readable medium storing instructions executable by a processor, wherein the instructions comprising instructions to:
capture, using a first optical device, first image data corresponding to a first view;
establish a first facial region within the first view;
determine a first cluster of facial feature points corresponding to the first facial region;
capture, using a second optical device, second image data corresponding to a second view;
establish a second facial region within the second view;
determine a second cluster of facial feature points corresponding to the second facial region;
compare the first cluster of facial feature points with the second cluster of facial feature points;
determine that the first cluster of facial feature points and the second cluster of facial feature points correspond to a rendered participant when a degree of similarity between the first cluster of facial feature points and the second cluster of facial feature points satisfies a first threshold; and
determine that the first cluster of facial feature points and the second cluster of facial feature points correspond to a genuine participant when the degree of similarity between the first cluster of facial feature points and the second cluster of facial feature points is below a second threshold.
18. The non-transitory computer readable medium of claim 17 , wherein the instructions further comprise instructions to render, using a display device coupled to the processor, third image data corresponding to the genuine participant, responsive to determining that the first cluster of facial feature points and the second cluster of facial feature points correspond to the genuine participant.
19. The non-transitory computer readable medium of claim 17 , wherein the instructions further comprise instructions to:
establish a rectangular polygon within the first view; and
determine that the first facial region is at least partially bounded by the rectangular polygon, and
wherein the instructions to determine the first cluster of facial feature points corresponding to the first facial region comprise instructions to determine the first cluster of facial feature points responsive to determining that the first facial region is at least partially bounded by the rectangular polygon.
20. The non-transitory computer readable medium of claim 17 , wherein the instructions to compare the first cluster of facial feature points with the second cluster of facial feature points comprise instructions to determine distances between two or more facial feature points of the first cluster of facial feature points and determine distances between two or more corresponding facial feature points within the second cluster of facial feature points.