Dynamic identity authentication
A method of identifying a person, the method comprising: acquiring spatiotemporal data for each of a plurality of anatomical landmarks associated with an activity engaged in by a person that defines a spatiotemporal trajectory of the anatomical landmark during the activity; modeling the acquired spatiotemporal data as a spatiotemporal graph (ST-Graph); and processing the ST-Graph using at least one non-local graph convolution neural network (NLGCN) to provide an identity for the person.
1. A method of identifying a person, the method comprising:
acquiring a sequence of video images of the person performing an activity;
identifying in the sequence of video images, anatomical landmarks that exhibit various degrees of motion associated with the activity engaged in by the person;
determining spatiotemporal trajectories exhibited by the identified anatomical landmarks in the sequence of video images;
constructing a spatiotemporal graph (ST-Graph) representing the spatiotemporal trajectories exhibited by the identified anatomical landmarks; and
processing the ST-Graph using at least one non-local graph convolution neural network (NLGCN) to provide an identity for the person.
2. The method according to claim 1 wherein the at least one NLGCN comprises at least one adaptive NLGCN (ANLGCN) including an adaptive adjacency matrix learned responsive to data relating anatomical landmarks of the identified anatomical landmarks that is not dictated solely by the person's physical body structure.
3. The method according to claim 1 , wherein each video frame includes an image of at least one body region of interest (BROI) imaging an anatomical landmark of the identified anatomical landmarks.
4. The method according to claim 3 and comprising processing the video frames to detect in each video frame the at least one BROI.
5. The method according to claim 3 and comprising identifying in each of the at least one detected BROI an image of an anatomical landmark of the identified anatomical landmarks.
6. The method according to claim 5 and comprising processing the images of the identified anatomical landmarks to determine the data defining the spatiotemporal trajectories.
7. The method according to claim 3 wherein the identified anatomical landmarks comprises joints.
8. The method according to claim 7 wherein the identified landmarks comprises bones connecting the joints.
9. The method according to claim 7 wherein the joints comprise finger knuckles.
10. The method according to claim 9 wherein the activity comprises a sequence of finger manipulations.
11. The method according to claim 10 wherein the finger manipulations comprise manipulations engaged in to operate a keyboard.
12. The method according to claim 7 wherein the joints comprise joints of the large appendages.
13. The method according to claim 12 wherein the activity is a sport.
14. The method according to claim 13 wherein the sport is soccer.
15. The method according to claim 13 wherein the sport is golf.
16. A method of identifying a person, the method comprising:
acquiring a sequence of video images of the person performing an activity;
identifying in the sequence of video images anatomical landmarks that exhibit various degrees of motion associated with the activity engaged in by the person;
determining spatiotemporal trajectories exhibited by the identified anatomical landmarks in the sequence of video images, wherein the identified anatomical landmarks comprises facial landmarks;
constructing a spatiotemporal graph (ST-Graph) representing the spatiotemporal trajectories exhibited by the identified anatomical landmarks; and
processing the ST-Graph using at least one non-local graph convolution neural network (NLGCN) to provide an identity for the person.
17. The method according to claim 16 wherein the facial landmarks comprise facial landmarks whose motions are used to define action units (AUs) of the facial action coding system (FACS) used to taxonomize facial expressions and micro-expressions.
18. The method according to claim 16 wherein the plurality of anatomical landmarks comprises minutia pair features of fingerprints of a plurality of fingers of a hand.