eprint.iacr.orgpaperGhostFaceMPC: Privacy-Preserving Face Recognition with Multi-Party Computation
GhostFaceMPC is presented as the first end-to-end face recognition system that runs a deep CNN feature extractor (GhostFaceNet) and the matching step under secure multi-party computation. Raw images and biometric embeddings are never exposed during extraction, storage or matching. The authors contribute an MPC-friendly quantization of GhostFaceNet and circuit optimizations that make the approach practical, and they release code at crypto-unict/GhostFaceMPC.
TL;DR First end-to-end MPC face recognition pipeline: quantized GhostFaceNet extraction and matching without exposing images or embeddings.
Also seen from another angle🧠 AI →
Organisations named here
- GhostFaceMPC
Technologies named here
- GhostFaceNet
Techniques named here
- MPC-friendly quantization of the GhostFaceNet model
Related in other
eprint.iacr.orgpaperRerandomizable Garbling for Arithmetic Circuits
aseemshrey.inarticle