LockEyeGaze

                               - Next-Gen Biometric Security

Your Eyes are more reliable than your Face

Why us? Why eyes?

LockEyeGaze revolutionizes authentication and deepfake detection using cutting-edge eye movement analysis. Our AI-powered system detects unique micro eye behaviors, creating secure, spoof-resistant authentication and unmatched deepfake detection.

Stronger Security, Smarter Authentication. Experience the future with LockEyeGaze.


Robust Authentication and Deepfake Detection 

We provide solution for governments, financial institutions, and cybersecurity firms, setting a new standard in digital identity protection and synthetic media defense.

Breakthrough Performance in Eye Movement-Based Authentication 

LockEyeGaze is the FIRST in the world to achieves over 90.39%[1] Precision (Recall 81.05%, AUC 0.935) at 30Hz and 60Hz, using only a standard RGB camera. Our system supports a wide range of eye-tracking techniques and devices, including gaze estimation-based eye-tracking algorithms as well as hardware solutions like Tobii and Pupil Labs eye trackers. This groundbreaking accuracy redefines gaze-based authentication, eliminating the need for specialized infrared or depth-sensing hardware. By leveraging micro eye movements and integrating our advanced face and gaze authentication layers, our system achieves an unprecedented overall accuracy of 99%+, delivering a seamless, highly secure, and cost-effective biometric security solution. 

Breakthrough Performance in Deepfake Detection 

LockEyeGaze achieves world-leading Precision over 99.92%[2] (Recall 96.21%, AUC 0.981) across Deepfake benchmarks by analyzing involuntary behaviours, which are nearly impossible to replicate in synthetic videos. Unlike traditional methods that rely on pixel inconsistencies, our system detects involuntary and micro behaviours, making it highly resistant to advanced AI-generated fakes. Designed for real-time, on-device processing, it ensures privacy and security without cloud dependency. 


[1] Data: 3 second time series eye movement data from 60hz RGB camera, Results on Pilot Test [2] Results on Public Datasets, Celeb-DF-v1 AUC=0.9894, Celeb-DF-v2 AUC=0.9722, UADFV AUC=0.9909

CyberSecurity Risks

Financial Security

Biometric Authentication

Need to speak with a specific expert from our team?

LinkedIn