LowKey: Leveraging Adversarial Attacks to Protect Social Media Users from Facial Recognition
This addresses privacy concerns for social media users against mass surveillance by corporations and governments, offering a practical tool for protection.
The authors tackled the problem of protecting social media users from facial recognition by developing an adversarial filter that is effective against industrial-grade systems, reducing the accuracy of Amazon Rekognition and Microsoft Azure Face Recognition API to below 1%.
Facial recognition systems are increasingly deployed by private corporations, government agencies, and contractors for consumer services and mass surveillance programs alike. These systems are typically built by scraping social media profiles for user images. Adversarial perturbations have been proposed for bypassing facial recognition systems. However, existing methods fail on full-scale systems and commercial APIs. We develop our own adversarial filter that accounts for the entire image processing pipeline and is demonstrably effective against industrial-grade pipelines that include face detection and large scale databases. Additionally, we release an easy-to-use webtool that significantly degrades the accuracy of Amazon Rekognition and the Microsoft Azure Face Recognition API, reducing the accuracy of each to below 1%.