03 / Open reference

FaceApp

Reference implementation for defence-in-depth biometric attendance

Rails · Flask · ONNX · 2025 · Academic

What it is

The working artefact behind my dissertation: roughly 3,000 lines of Ruby and Python demonstrating that an eight-layer attendance system — face verification, active and passive liveness, geofencing and an HMAC-bound location token — runs on commodity hardware with no GPU. Published with a full threat model, attack tree and a candid residual-risk analysis.

8
Verification layers
~3k
Lines of Ruby + Python
130
References, all ≥2019
Talk about it
Contact

Let's build
something honest.