4

I am looking to use AWS Reckognition in one of my projects and trying to find out whether or not its possible to differentiate between a still image (photograph) vs a real person, in other words liveness detection. I don't want my system to be fooled with a still photograph for authentication.

I see that it has many features such as pose and emotion detection, etc. If its not an official feature, is there a work around or any tricks that some of you have used to achieve what I want?

I am also wondering if its possible to detect gaze and how to best approach that. I want to see where the user is looking at, at the screen, to the side, etc.

Alternatively, if AWS does not have a good solution for this, what are some of your alternative recommendations?

Regards

prosonic
  • 53
  • 1
  • 5
  • I don't think its a simple solution as you think. It requires understanding of 3d data. Try to procure some stereo depth Cameras and understand it. After that train your data. – vishnukumar Jun 21 '20 at 15:50
  • 1
    Does this answer your question? [AWS live human detection from video](https://stackoverflow.com/questions/60541185/aws-live-human-detection-from-video) – Hugo Jun 24 '20 at 03:10
  • Hey, were you able to solve this problem? I try to achieve the same. If yes, how did you do it? Through AWS or OpenCV? An article I found for reference: https://aws.amazon.com/blogs/industries/improving-fraud-prevention-in-financial-institutions-by-building-a-liveness-detection-solution/ – anshaj Sep 22 '20 at 15:12
  • I didn't use OpenCV. I just used AWS and compared landmark positions between two images, position of eyes/mouth. Its not perfect but okay enough for my needs. – prosonic Sep 22 '20 at 15:50
  • @prosonic, could you please elaborate on your solution? – Opal Sep 24 '20 at 10:28
  • Yes take two pictures about 200-500 ms apart, and start comparing the position of important landmarks (eyes, nose, mouth), use that to compute a score. If there is too little difference, means its still. If there are some differences, u define ur threshold for whats live and whats not. – prosonic Sep 25 '20 at 19:07

3 Answers3

2

Could you make use of blink detection, which isnt part of AWS Rekognition, to check if an image isn't a still photograph. You just need OpenCV.

Here is an example.

Face recognition alone is notoriously insecure when it comes to authentication, as has been evidenced by the many examples of the Android Face Unlock functionality being fooled by photographs.

Apple makes use of Depth sensing cameras in its FaceID technology to create a 3D map of the faces which cant be fooled by a photograph. Windows Hello face authentication utilises a camera specially configured for near infrared (IR) imaging to authenticate.

Opal
  • 81,889
  • 28
  • 189
  • 210
George
  • 21
  • 2
1

as alternative to gaze you can have a look at liveness example using aws rekognition based on face and nose position:

dmitrdv
  • 61
  • 3
0

AWS just launched new Rekognition liveness support which can be used - https://docs.aws.amazon.com/rekognition/latest/dg/face-liveness.html

Blog - https://aws.amazon.com/blogs/machine-learning/detect-real-and-live-users-and-deter-bad-actors-using-amazon-rekognition-face-liveness/

rishabhjainps
  • 410
  • 1
  • 5
  • 14