INSIGHTFACE SERVER
Dashboard
LOCAL FACE INTELLIGENCE
Your face recognition service, at a glance.
Detect, check liveness, compare, enroll, and search without sending images outside your server.
Execution provider
Checking…
SERVICE
Runtime status
THIS TAB
Recent errors
No recent errorsAPI failures from this browser tab will appear here.
IDENTITY SPACES
Collections
Each collection is pinned to the model and preprocessing version active when it is created.
| Collection | Threshold | Profile | Crops | Model | People | Faces | Actions |
|---|
No collections yetCreate one to start enrolling people.
ENROLLMENT
People & face samples
Register several clear photos for each real person to improve search robustness.
DIRECTORY
People
Select a collectionIts enrolled people will appear here.
No person selectedChoose a person to view details and face samples.
STATELESS
Detect faces
Find faces, five-point landmarks, confidence, and quality in one image.
Image preview
ONE-TO-ONE
Compare two faces
The selected detection profile chooses one face from each image for cosine comparison.
ONE-TO-MANY
Search a collection
Find the closest enrolled people using the Collection single-face selection strategy.
Query preview
RTSP MONITORS
Camera monitoring
Create persistent server-side Monitors, inspect their current state, and pull recent in-memory events. Video frames and events are not saved.
CONFIGURATION
Monitors
No MonitorsCreate a Monitor to connect an RTSP camera.
No Monitor selectedSelect a Monitor to inspect its live recognition state.
Recognized person
Detected, not enrolled
CURRENT STATE
Select a Monitor
- Source
- —
- Collection
- —
- Match threshold
- —
- Recent event capacity
- —
Select a Monitor to inspect it.
- Connection
- —
- Effective rate
- —
- Inference time
- —
- Skipped frames
- —
IN-MEMORY BUFFER
Recent events
ADMINISTRATION
System
Runtime diagnostics and safe system configuration. Secrets and embeddings are never displayed.
Liveness
Liveness status unknownDownload the liveness model and save its startup setting here. Enabling it requires a manual Server restart.
Install the addon with the model download command, add liveness to inference.addons in the existing startup configuration, then restart Server.
docker compose -f server/deploy/compose.cpu.yml run --rm models addons install livenessUse the Compose file for your deployment.
RUNTIME
Hardware & software
STORAGE
Database & mounts
MODELS
Loaded models
DEVELOPER TOOLS
OpenAPI Schema
Inspect the exact paths, parameters, and response schemas exposed by this running server.
DOCUMENTATION
Help
Beginner guidance and complete API documentation are bundled with this server and work without internet access.
Loading documentation…