Skip to main content

How Facial Recognition Works

SnapdMe uses facial recognition to match your selfie against faces in the event photo gallery. Here's what happens under the hood.

The matching process

  1. Selfie capture — you take a selfie through your browser.
  2. Feature extraction — SnapdMe extracts a numerical representation of your face (a "facial signature"). This is not a stored image — it is a set of measurements.
  3. Gallery scan — your facial signature is compared against the signatures of every face detected in the event photos.
  4. Results — photos where your face is detected above a confidence threshold are shown to you.

Accuracy

Facial recognition works best when:

  • The selfie is taken in good lighting.
  • Your face is unobstructed (no heavy glasses, hats, or masks).
  • The event photos are reasonably clear (blurry or heavily cropped faces may not match).

If you believe a photo of you is missing, try retaking your selfie in better light, or browse the full gallery manually.

Privacy and data

  • Your selfie is processed in real time and not stored after matching is complete.
  • Facial signatures extracted from event photos are deleted when the event is archived or deleted by the organizer.
  • SnapdMe does not use your facial data for any purpose other than finding your event photos.