CipherFace: A Fully Homomorphic Encryption-Driven Framework for Secure Cloud-Based Facial Recognition
Sefik Serengil, Alper Ozpinar · 2025
Read CipherFace: A Fully Homomorphic Encryption-Driven Framework for Secure Cloud-Based Facial Recognition by Sefik Serengil, Alper Ozpinar (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-18514.txt.
Description
Sefik Serengil and Alper Ozpinar move face recognition into the cloud without ever decrypting a face, in a 2025 paper. Systems store face embeddings, compact numeric portraits, and check identity by distance between them. Those portraits still leak private information, and ordinary encryption blocks the math needed to compare them.
Their framework, CipherFace, computes distances while everything stays encrypted under fully homomorphic encryption. New encrypted formulas cover both straight-line and angle-based distances. Tests across face models, portrait sizes, and security settings showed the approach scales. The open-source release lets hospitals, banks, and border systems outsource matching without handing over faces. The cloud does the math, yet it never sees a single face in the clear. That split is the whole point of the design.
Catalog record
| CREATOR | Sefik Serengil, Alper Ozpinar |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 30 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-18514