Automated detection of unauthorized access to device screens using machine learning facial recognition techniques
Methods, apparatus, and processor-readable storage media for automated detection of unauthorized access to device screens using machine learning facial recognition techniques are provided herein. An example computer-implemented method includes automatically detecting, using one or more machine learning facial recognition techniques, at least a portion of at least one face within a given proximity of a user device; automatically performing a comparison, using one or more machine learning facial recognition techniques, of the detected portion of the at least one face with facial images attributed to a set of trusted individuals; and automatically performing security-related actions with respect to the user device upon determining, based at least in part on the comparison, that the detected portion of the at least one face exhibits at least a threshold level of distinctiveness relative to each of the facial images attributed to the set of trusted individuals.
1. A computer-implemented method comprising:
automatically detecting, using one or more machine learning facial recognition techniques, at least a portion of at least one face within a given proximity of a user device;
automatically performing a comparison, using one or more machine learning facial recognition techniques, of the detected portion of the at least one face with one or more facial images attributed to a set of one or more trusted individuals;
automatically performing one or more security-related actions with respect to the user device upon determining, based at least in part on the comparison, that the detected portion of the at least one face exhibits at least a threshold level of distinctiveness relative to each of the one or more facial images attributed to the set of one or more trusted individuals; and
automatically reversing at least a portion of the one or more performed security-related actions with respect to the user device upon failing to detect, for a predetermined length of time, at least a portion of at least one face within the given proximity of the user device that exhibits at least the threshold level of distinctiveness relative to each of the one or more facial images attributed to the set of one or more trusted individuals;
wherein the method is performed by at least one processing device comprising a processor coupled to a memory.
2. The computer-implemented method of claim 1 , wherein automatically detecting comprises automatically detecting at least a portion of at least one face upon mapping, using the one or more machine learning facial recognition techniques, at least a user-configurable percentage of the at least one face.
3. The computer-implemented method of claim 1 , wherein automatically detecting comprises automatically detecting at least a portion of at least one face upon observing at least a portion of the at least one face for at least a user-configurable amount of time.
4. The computer-implemented method of claim 1 , wherein the one or more machine learning facial recognition techniques comprise at least one Eigenface-based technique.
5. The computer-implemented method of claim 1 , wherein the one or more machine learning facial recognition techniques used in automatically detecting the at least a portion of at least one face within the given proximity of the user device differ from the one or more machine learning facial recognition techniques used in automatically performing the comparison.
6. The computer-implemented method of claim 1 , wherein the one or more machine learning facial recognition techniques used in automatically detecting the at least a portion of at least one face within the given proximity of the user device comprise the same one or more machine learning facial recognition techniques used in automatically performing the comparison.
7. The computer-implemented method of claim 1 , wherein automatically performing the one or more security-related actions comprises generating a secure view and displaying the secure view via a screen of the user device.
8. The computer-implemented method of claim 7 , wherein generating the secure view comprises deactivating one or more user-configurable applications.
9. The computer-implemented method of claim 7 , wherein generating the secure view comprises enabling execution of one or more user-configurable applications.
10. The computer-implemented method of claim 7 , wherein generating the secure view comprises enabling modified execution of one or more user-configurable applications in accordance with one or more user instructions.
11. The computer-implemented method of claim 7 , further comprising:
deactivating the secure view and removing the secure view from the screen of the user device in response to one or more user actions.
12. The computer-implemented method of claim 1 , wherein automatically performing the one or more security-related actions comprises at least one of pausing one or more audio-visual applications executing on the user device and stopping one or more audio-visual applications executing on the user device.
13. The computer-implemented method of claim 1 , wherein automatically performing the one or more security-related actions comprises at least one of pausing one or more audio applications executing on the user device and stopping one or more audio applications executing on the user device.
14. The computer-implemented method of claim 1 , further comprising:
modifying the set of one or more trusted individuals based at least in part on user input.
15. A non-transitory processor-readable storage medium having stored therein program code of one or more software programs, wherein the program code when executed by at least one processing device causes the at least one processing device:
to automatically detect, using one or more machine learning facial recognition techniques, at least a portion of at least one face within a given proximity of a user device;
to automatically perform a comparison, using one or more machine learning facial recognition techniques, of the detected portion of the at least one face with one or more facial images attributed to a set of one or more trusted individuals;
to automatically perform one or more security-related actions with respect to the user device upon determining, based at least in part on the comparison, that the detected portion of the at least one face exhibits at least a threshold level of distinctiveness relative to each of the one or more facial images attributed to the set of one or more trusted individuals; and
to automatically reverse at least a portion of the one or more performed security-related actions with respect to the user device upon failing to detect, for a predetermined length of time, at least a portion of at least one face within the given proximity of the user device that exhibits at least the threshold level of distinctiveness relative to each of the one or more facial images attributed to the set of one or more trusted individuals.
16. The non-transitory processor-readable storage medium of claim 15 , wherein automatically detecting comprises automatically detecting at least a portion of at least one face upon mapping, using the one or more machine learning facial recognition techniques, at least a user-configurable percentage of the at least one face.
17. The non-transitory processor-readable storage medium of claim 15 , wherein automatically detecting comprises automatically detecting at least a portion of at least one face upon observing at least a portion of the at least one face for at least a user-configurable amount of time.
18. An apparatus comprising:
at least one processing device comprising a processor coupled to a memory;
the at least one processing device being configured:
to automatically detect, using one or more machine learning facial recognition techniques, at least a portion of at least one face within a given proximity of a user device;
to automatically perform a comparison, using one or more machine learning facial recognition techniques, of the detected portion of the at least one face with one or more facial images attributed to a set of one or more trusted individuals;
to automatically perform one or more security-related actions with respect to the user device upon determining, based at least in part on the comparison, that the detected portion of the at least one face exhibits at least a threshold level of distinctiveness relative to each of the one or more facial images attributed to the set of one or more trusted individuals; and
to automatically reverse at least a portion of the one or more performed security-related actions with respect to the user device upon failing to detect, for a predetermined length of time, at least a portion of at least one face within the given proximity of the user device that exhibits at least the threshold level of distinctiveness relative to each of the one or more facial images attributed to the set of one or more trusted individuals.
19. The apparatus of claim 18 , wherein automatically detecting comprises automatically detecting at least a portion of at least one face upon mapping, using the one or more machine learning facial recognition techniques, at least a user-configurable percentage of the at least one face.
20. The apparatus of claim 18 , wherein automatically detecting comprises automatically detecting at least a portion of at least one face upon observing at least a portion of the at least one face for at least a user-configurable amount of time.