FunshadeRS: Rust Implementation of the Privacy-Preserving Funshade Framework for Distributed Systems

With biometric data increasingly being used for security, protecting this sensitive information is more important than ever. As a particular example, the Digidow project uses biometric authentication for real-world transactions such as opening doors. Digidow employs a decentralised structure in which the biometric template of each person is stored either with a chosen cloud provider or on a personal server. Since biometric data are stored and processed across potentially untrusted locations, comparison of the stored template with live data from a sensor requires a secure and privacy-preserving solution that protects this sensitive data even in the presence of potentially malicious parties.
Funshade provides a method to compare biometric data between parties without revealing the actual data itself, ensuring privacy and security. It is designed to enable secure biometric authentication through privacy-preserving protocols. In contrast to previous k-of-n bitwise comparison approaches, Funshade supports full threshold comparison over usual vector metrics, and is therefore applicable to state-of-the-art face embedding models. We implement a complete open source prototype in Rust, chosen for its strong memory-safety and performance features, to bridge the gap between modern face recognition pipelines and multi-party computation for decentralised architectures.