FaceGuard: Enhancing Data Privacy through Selective Facial Anonymization in Public Events

Personal data protection, mandated by regulations such as GDPR and PDPA, necessitates ethical handling of personal information. A common issue arises with the unintentional publication of facial images, which can occur in group photos at seminars, on social media, or in various other scenarios. An effective solution is to anonymize these images either upon request or proactively to maintain privacy and compliance. This research proposes an online system for automatic face detection and anonymizing or blurring specifically targeted unknown persons (never previously recognized). The system provides three main services: (1) selectively blurring faces of non-consenting individuals in seminar photos using identifiable markers like special lanyards, (2) blurring all faces except the user’s in personal social media posts, and (3) offering semi-automatic and manual blurring options for general applications. This system aims to enhance privacy protection by ensuring that individuals’ facial data is handled responsibly and in accordance with legal requirements. It emphasizes a seamless processing experience, prioritizing ease of use while striking a balance between robust security measures and maintaining user privacy. This work-in-progress paper does not focus on enhancing accuracy levels but instead proposes a framework designed to differentiate and anonymize targeted unknown persons, aligning with personal data protection compliance.