Face recognition and liveness through an API.
A proof of concept for integrating biometric checks into applications.
LiveXFace is an API-first project for face recognition and liveness detection. The goal is to keep application-facing APIs separate from model inference and face search, so each part can evolve without changing the way applications integrate.
| Area | Focus |
|---|---|
| Face recognition | Match faces and search stored embeddings. |
| Liveness | Assess whether a face capture comes from a live person. |
| Integration | Expose these capabilities through clear API boundaries. |
The POC is being developed around the following components:
| Component | Technology |
|---|---|
| API | Go, Gin |
| Inference | Python, Flask, InsightFace |
| Face search and storage | PostgreSQL, pgvector |
| Cache and messaging | Redis, NATS |
| Routing | Traefik |
LiveXFace is in the POC stage. Interfaces, documentation, and repository structure may change as the product is validated. This README describes the project's direction; it is not a production-readiness or accuracy claim.