Safeguarding Identities with GAN-based Face Anonymization

Abstract

Effective anonymous facial registration techniques are critical to address privacy concerns arising from facial recognition technology. This study presents an intelligent anonymity platform that incorporates blockchain with advanced privacy and uses a CIAGAN-powered approach. This solution addresses the immediate need for privacy in facial recognition technology. The proposed system uses advanced techniques to anonymously generate highly realistic and effective facial images. The widespread use of facial recognition systems places greater emphasis on privacy concerns, emphasizing the need for strong enrollment mechanisms. The proposed system uses CIAGAN to address this challenge and generate facial images while preserving important attributes. Blockchain storage ensures that data integrity and security are maintained. The process begins with detailed image preprocessing steps to improve data quality and eliminate unwanted noise. CIAGAN can generate anonymous face images with important facial attributes to complicate the recognition of specific objects. A dataset of 202,599 facial images was used. Performance metrics such as PSNR and SSIM indicate image quality and uniformity. The PSNR obtained was 35.0516, indicating a unique image anonymization process.