How a73720177-design/facr_recognize works
This repository implements a local, camera-based two-factor safe: a user selects one of three numbered compartments, passes face identification, then enrolls or verifies an ordered hand/arm-pose pattern. Face registration is a distinct setup stage; ownership and reset rules govern compartment access, while successful pattern verification unlocks the software state only (no physical lock is connected). Edges reflect the sampled callers and README-documented workflow; some storage and menu relationships are documented rather than shown in complete source excerpts.
Main entry points: safe_app.py Safe session UI · face_identity.py Face identification · recognition.py Menu and CLI · vision_utils.py Pose and hand detectors
Components 13
Entry and UI
- Menu and CLI
recognition.py - Safe session UI
safe_app.py
Face authentication
- Face identification
face_identity.py - Face enrollment data
- Face stability gate
face_gate.py
Gesture authentication
- Pose and hand detectors
vision_utils.py - Pose descriptors
safe_pattern.py - Pattern enrollment and check
safe_pattern.py
Ownership and state
- Compartment ownership
safe_slot.py - Compartment records
Other components
- Safe user
- Camera feed
- Model hosting
Connections 18
- Safe user → Menu and CLI: selects task
README.md - Menu and CLI → Safe session UI: starts session
recognition.py - Safe user → Camera feed: presents face and pose
README.md - Camera feed → Safe session UI: supplies frames
safe_app.py - Menu and CLI → Face identification: enrolls face
recognition.py - Face identification → Face enrollment data: reads and writes
face_identity.py - Safe session UI → Face identification: identifies face
safe_app.py - Safe session UI → Face stability gate: confirms identity
safe_app.py - Safe session UI → Pose and hand detectors: creates detectors
safe_app.py - Pose and hand detectors → Pose descriptors: provides landmarks
safe_app.py - Safe session UI → Pose descriptors: extracts pose
safe_app.py - Safe session UI → Pattern enrollment and check: updates pattern
safe_app.py - Pattern enrollment and check → Compartment ownership: saves samples
safe_pattern.py - Safe session UI → Compartment ownership: checks ownership
safe_app.py - Compartment ownership → Compartment records: reads and writes
README.md - Face identification → Model hosting: downloads face models
face_identity.py - Pose and hand detectors → Model hosting: downloads pose models
vision_utils.py - Safe session UI → Safe user: displays status
safe_app.py