biofrq.liveness

Passive presentation-attack detection. One image in, a live / spoof verdict out plus a confidence score.

Params

Field	Type	Required	Default	Notes
`images[0]`	`ImageItem`	yes	—	One image. See Envelope → Image inputs.
`livenessThreshold`	float	no	tenant default	0–1. Higher = stricter. Only moves the verdict; the raw `score` is unaffected.

Request

json

{
  "action": "biofrq.liveness",
  "params": {
    "images": [
      { "kind": "image", "bytes": "<base64>", "contentType": "image/jpeg" }
    ],
    "livenessThreshold": 0.5
  }
}

Response

json

{
  "ok": true,
  "requestId": "...",
  "action": "biofrq.liveness",
  "data": {
    "results": [
      {
        "index": 0,
        "correlationId": "auto-0",
        "ok": true,
        "value": {
          "verdict": "live",
          "score": 0.94,
          "reasons": []
        }
      }
    ]
  },
  "errors": []
}

reasons is populated with short tags when verdict is "spoof" (e.g. "screen", "print", "noFace").

Common errors

INVALID_REQUEST — no face in the image, or livenessThreshold out of range.

Open in playgroundbiofrq.liveness