We only need to process one single image to retrieve facial vectors. After processing, the image will be deleted.
Only facial vectors will be stored in a decentralized and heavily encrypted database.
Our Glimpse recognition will work in poor light conditions, extreme pose variations, bad resolution, all facial expressions and partly occluded faces.
The recognition engine is robust to:
- Partial face occlusion
- Pose variations (± 45° deviation from frontal image)
- Gesture changes
- Beard and hairstyle changes
- Glasses (including sunglasses)
- Lighting changes
- Age variation (