Liveness Detection

Confirm real presence. Block every spoofing attack.

ASTERIA KYC biometric liveness detection determines whether the person submitting the selfie is physically present. Passive analysis, active challenge response, and injection attack detection run in sequence to return a single structured liveness result.

ASTERIA KYCLiveness Detection Result
PASSED
Method Usedpassive
Attack Typenone
liveness_resultpassed
confidence_score96
challenge_resultnot_triggered
injection_signalnone
Confidence96
Session IDakyc_lv_e3a9c2f1
Liveness detection

A real-person check that resists screen replays, printed photos, and injection attacks.

ASTERIA KYC liveness check — face capture with anti-spoofing signals and real-time guidance
3Detection Layers
6Attack Types
<500msResult Return
99%+Accuracy
Detection Methods

Three layers of detection. One liveness result.

Passive analysis runs first on every session. Challenge and injection layers activate where the passive signal warrants escalation.

Passive Analysis

Frame-level biometric signals analyzed without user interaction. Texture, depth cues, and micro-movement patterns assessed in the submitted capture.

Active Challenge

Randomised challenge sequence presented when passive signals are uncertain. Head movement, blink, or expression prompts configured per session.

Injection Attack Detection

Virtual camera, emulated device, and media injection signals detected at the session layer before biometric analysis begins.

Real-time signals

Face match, anti-spoof signals, and capture guidance — every step instrumented for audit-ready evidence.

ASTERIA KYC liveness check with callouts — face matched, liveness passed, hold-still capture guidance, and verification-in-progress signals
Attack Coverage

Every known spoofing vector. Covered.

Liveness checks are only meaningful if they cover the attack types that reach production. ASTERIA KYC covers the full range of known presentation attack categories.

01

Printed Photo

Static image printed or displayed on a second screen. Detected through texture depth analysis and micro-movement signal absence.

02

Video Replay

Pre-recorded video of the real user replayed during the session. Motion pattern analysis and frame consistency checks applied.

03

3D Mask

Silicone, paper, or rigid mask worn over the face. Surface texture analysis and depth signal inconsistency detection applied.

04

Deepfake Stream

AI-generated face substituted in real time during capture. Frequency artifacts, temporal consistency, and face boundary signals assessed.

05

Camera Injection

Virtual camera driver or media injection tool substituting the real device input. Session-layer device signal validation applied.

06

Partial Spoof

Photo or mask covering only part of the face. Face completeness validation and zone-level signal consistency checks applied.

Detection Outputs

Structured liveness result with confidence and reason codes.

Every liveness session returns a result code, confidence band, detected attack type where applicable, and reason codes for downstream use.

FieldDescriptionValues
liveness_resultOverall liveness decisionpassed / uncertain / failed
confidence_scoreLiveness confidence0–100
method_usedDetection method appliedpassive / challenge / injection
attack_typeDetected attack categorynone / print / replay / mask / deepfake / injection
challenge_resultActive challenge outcomepassed / failed / not_triggered
reason_codesStructured reason arrayarray of strings
Sample Output
{
  "liveness_result": "passed",
  "confidence_score": 96,
  "method_used": "passive",
  "attack_type": "none",
  "challenge_result": "not_triggered",
  "reason_codes": ["LIVENESS_PASSIVE_CLEAR"]
}
Result States

Liveness outcome built for operational escalation.

Uncertain liveness results do not automatically fail the session. The result feeds into the risk score and may trigger human review based on configuration.

passed

Passed

Passive or challenge analysis confirmed real presence. No spoofing signals detected. Confidence score above the configured threshold.

uncertain

Uncertain

One or more signals returned low confidence. Image quality, motion pattern, or challenge response could not be resolved to a passing threshold.

failed

Failed

A spoofing signal was detected above the failure threshold. Attack type recorded in the result. Session escalated or rejected based on platform configuration.

Ready to turn identity verification into a controlled compliance workflow?

Use ASTERIA KYC to verify users, screen risk, and preserve evidence from one connected platform.

ASTERIA KYC | Compliance-First Identity Verification Infrastructure