Authentication techniques including speech and/or lip movement analysis

1. A method comprising:

receiving a request to authenticate a user;

presenting one or more screen layouts to the user;

capturing a sequence of images which include the user's eyes as the one or more screen layouts are displayed;

performing eye movement detection across the sequence of images to identify a correlation between motion of the user's eyes as the one or more screen layouts are presented and an expected motion of the user's eyes as the one or more screen layouts are presented;

capturing audio of the user's voice;

performing voice recognition techniques to determine a correlation between the captured audio of the user's voice and one or more voice prints;

training a lip movement analysis module to associate particular phonetics and volume levels with particular lip or mouth positions and/or movements over time;

performing lip-movement analysis using the trained lip movement analysis module to determine a correlation between the audio of the user's voice and detected motion of the user's lips across the sequence of images;

detecting motion of the user's lips across the sequence of images;

correlating the lip motion with a reference lip motion captured during an enrollment of the user;

and

generating an assurance level based on (a) a correlation between the images of the user's face and facial template data associated with the user, (b) a correlation between the motion of the user's eyes and an expected motion of the user's eyes as the one or more screen layouts are presented, (c) a correlation between the captured audio of the user's voice and the one or more voice prints, and (d) a correlation between the audio of the user's voice and detected motion of the user's lips across the sequence of images.

2. The method as in claim 1 further comprising:

performing facial recognition to identify the correlation between one or more images of the user's face and facial template data associated with the user.

3. The method as in claim 2 wherein presenting the one or more screen layouts comprises displaying one or more graphics images, photographs, or motion video images in designated regions of the display, wherein the expected motion characteristic of the user's eyes depends on the contents, the size, the appearance, the context and the region where these elements are shown.

4. The method as in claim 3 wherein facial recognition is performed first, resulting in a first correlation and eye movement detection is performed only if the facial recognition results in a correlation above a specified threshold.

5. The method as in claim 3 wherein the designated regions of the display for displaying the graphics images, photographs, or motion video images are selected randomly.

6. The method as in claim 3 wherein presenting the one or more screen layouts comprises displaying a plurality of graphics images, photographs, or motion video images in a specified sequence using a specified timing.

7. The method as in claim 1 further comprising: determining whether the assurance level is sufficient to authenticate the user for a particular transaction; and permitting the transaction if the assurance level is sufficient.

8. The method as in claim 7 wherein the transaction is an online transaction with a remote relying party.

9. The method as in claim 7 further comprising:

if the assurance level is not sufficient by itself to authenticate the user for a particular transaction, then combining the assurance level with authentication data generated using one or more additional authentication techniques to authenticate the user for the particular transaction.

10. The method as in claim 9 wherein one of the additional authentication techniques comprises performing an explicit authentication in which the end user is required to enter a password or PIN or to swipe a finger on a fingerprint authentication device.

11. The method as in claim 9 wherein one of the additional authentication techniques comprises one or more non-intrusive authentication techniques in which sensor data collected from one or more device sensors is used for authentication.

12. The method as in claim 11 wherein the sensor data comprises location data.

13. The method as in claim 11 wherein one of the non-intrusive techniques comprises determining elapsed time from the last explicit user authentication and responsively increasing the assurance level of explicit user authentication has occurred within a specified time period.

14. The method as in claim 1 wherein the expected motion of the user's eyes is based on learned characteristics of eye-movement triggered by different types of images.

15. The method as in claim 1 wherein capturing audio of the user's voice comprises prompting the user to speak one or more words and/or phrases and receiving responses from the user via a microphone.

16. An apparatus comprising:

an authentication engine receiving a request to authenticate a user;

a camera capturing a sequence of images which include the user's eyes as one or more screen layouts are displayed;

an eye tracking module performing eye movement detection across the sequence of images to identify a correlation between motion of the user's eyes as the one or more screen layouts are presented and an expected motion of the user's eyes as the one or more screen layouts are presented;

a voice recognition module to perform voice recognition techniques to determine a correlation between captured audio of the user's voice and one or more voice prints; and

a lip movement analysis module, trained to associate particular phonetics and volume levels with particular lip or mouth positions and/or movements over time, to perform lip-movement analysis to determine a correlation between the audio of the user's voice and detected motion of the user's lips across the sequence of images, the lip movement analysis module further configured to detect motion of the user's lips across a sequence of images and correlate the lip motion with a reference lip motion captured during an enrollment of the user; and

the authentication engine generating an assurance level based on (a) a correlation between the images of the user's face and facial template data associated with the user, (b) the correlation between the motion of the user's eyes and an expected motion of the user's eyes as the one or more screen layouts are presented; and (c) the correlation between the captured audio of the user's voice and the one or more voice prints, and (d) the correlation between the audio of the user's voice and detected motion of the user's lips across the sequence of images.

17. The apparatus as in claim 16 further comprising:

a facial recognition engine to perform facial recognition to identify the correlation between one or more images of the user's face and facial template data associated with the user.

18. The apparatus as in claim 17 wherein presenting the one or more screen layouts comprises displaying one or more graphics images, photographs, or motion video images in designated regions of the display, wherein the expected motion of the user's eyes is in the direction of the graphics images, photographs, or motion video images to different degrees.

19. The apparatus as in claim 18 wherein facial recognition is performed first, resulting in a first correlation and eye movement detection is performed only if the facial recognition results in a correlation above a specified threshold.

20. The apparatus as in claim 18 wherein the designated regions of the display for displaying the graphics images, photographs, or motion video images are selected randomly.

21. The apparatus as in claim 18 wherein presenting the one or more screen layouts comprises displaying a plurality of graphics images, photographs, or motion video images in a specified sequence using a specified timing.

22. The apparatus as in claim 21 wherein the transaction is an online transaction with a remote relying party.

23. The apparatus as in claim 16 further comprising: the authentication engine determining whether the assurance level is sufficient to authenticate the user for a particular transaction; and permitting the transaction if the assurance level is sufficient.

24. The apparatus as in claim 23 further comprising:

the authentication engine combining the assurance level with authentication data generated using one or more additional authentication techniques to authenticate the user for the particular transaction if the assurance level is not sufficient by itself to authenticate the user for a particular transaction.

25. The apparatus as in claim 23 wherein one of the additional authentication techniques comprises one or more non-intrusive authentication techniques in which sensor data collected from one or more device sensors is used for authentication.

26. The apparatus as in claim 25 wherein one of the additional authentication techniques comprises performing an explicit authentication in which the end user is required to enter a password or PIN or to swipe a finger on a fingerprint authentication device.

27. The apparatus as in claim 25 wherein the sensor data comprises location data.

28. The apparatus as in claim 25 wherein one of the non-intrusive techniques comprises determining elapsed time from the last explicit user authentication and responsively increasing the assurance level of explicit user authentication has occurred within a specified time period.

29. The apparatus as in claim 16 wherein the expected motion of the user's eyes is based on learned characteristics of eye-movement triggered by different types of images.

30. The apparatus as in claim 16 wherein capturing audio of the user's voice comprises prompting the user to speak one or more words and/or phrases and receiving responses from the user via a microphone.