face-recognition-pipeline

What problem does it solve?

This Skill centralizes and enforces a robust facial recognition check-in pipeline for event totems and backend systems, reducing false positives, preventing duplicate entries, and ensuring adherence to event AI configuration.

Core Features & Use Cases

• Capture-to-Checkin Flow: Coordinates face capture, embedding generation, similarity comparison, and final check-in registration.
• Confidence & Validation: Applies configurable confidence thresholds from EventAIConfig and rejects low-confidence matches.
• Safety Guards: Prevents duplicate check-ins and enforces cooldown windows to mitigate replay and rapid-repeat attempts.
• Use Case: Validate an attendee at a kiosk using a face capture, match against stored embeddings, and register a single, auditable check-in or fallback to alternative methods when needed.

Quick Start

Use the face-recognition-pipeline to validate a captured face against stored attendee embeddings and register a check-in when the confidence exceeds the event threshold.