We are creative, ambitious and ready for challenges! Hire Us
SCALABLE FACE RECOGNITION WITH ANTI-SPOOFING
Client’s Need
Global $ 800 bln auto insurance industry heavily depends on manual vehicle dent inspections. This method is labor-intensive and can be inconsistent due to human variability. Client desired an automated and objective solution for dent assessments and repair estimations.
Objectives
- Develop face recognition capable of anti-spoofing.
- Compatibility with all single front facing camera smart phones.
- Under 5% false positives in face detection and recognition in the dataset shared.
Our Approach
- Solution compatible with both Android and iOS.
- Face Detection and recognition with primitive front facing camera.
Implementation of ‘Challenge – Response Anti-Spoofing‘ with inherent less than 1% cheat probability*
Result
- Performance: Achieved 13 fps** of detection, recognition and anti spoofing processing.
- 0% false positive achieved, ensuring robust security without specialized hardware.
Scalable solution, additional gestures and factors can be added considering hardware and human interface fit on application target market.
Published:
September 28, 2024
Category:
Edge Computing
Client:
Oceanthemes
