Your passport to the 7th continent
summary
Mobile devices use the built‑in camera to enroll users by scanning an ID photo plus a live face and storing a face print (claimed 80 nodal points, with 15–20 often sufficient). Liveness is enforced with gesture sequences, head‑tracking targets and an illumination color challenge that detects screen vs skin, plus pulse/eye‑region checks; these reduce replay/video attacks but silicone‑mask proof next year remains an open problem. Systems tune an acceptance threshold to trade off false rejection vs false acceptance, causing measurable user friction.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 3:40Epic Love Story
- 13:41The sequel
- 15:46Spoofing
- 16:36Verification Score
- 18:11User Data
- 21:05Use Case
- 22:29Face Capture
- 25:00Spoofer
- 28:08Spoof
- 29:51Illumination detection
- 33:09Market expectations
- 35:22Lymans challenge
- 35:43Laser challenge
- 36:47Explanation
- 37:59Innovation challenge
related talks
26mVirtual U: Defeating Face Liveness Detection by Building Virtual Models...
22mThe high cost of cheap biometrics
16mYour FACE in a SPACE
28mML solutions for the real world problems of identity verification
12mAre Users Able to Take Control of Their Data?
08mAdvancements in facial recognition
25mBiometric Authentication Under Threat: Liveness Detection Hacking
37mThe Living Dead: Hacking Mobile Face Recognition SDKs with Non-Deepfake Attacks