https://github.com/credona/age-decision-antispoof
Image-based anti-spoofing service detecting presentation attacks using ONNX models and multi-signal heuristics.
https://github.com/credona/age-decision-antispoof
anti-spoofing api biometrics computer-vision credona face-analysis face-detection fastapi liveness-detection machine-learning onnx open-source presentation-attack-detection privacy security
Last synced: about 1 month ago
JSON representation
Image-based anti-spoofing service detecting presentation attacks using ONNX models and multi-signal heuristics.
- Host: GitHub
- URL: https://github.com/credona/age-decision-antispoof
- Owner: credona
- License: apache-2.0
- Created: 2026-04-26T00:53:31.000Z (about 2 months ago)
- Default Branch: main
- Last Pushed: 2026-04-28T05:28:49.000Z (about 2 months ago)
- Last Synced: 2026-04-28T07:34:32.297Z (about 2 months ago)
- Topics: anti-spoofing, api, biometrics, computer-vision, credona, face-analysis, face-detection, fastapi, liveness-detection, machine-learning, onnx, open-source, presentation-attack-detection, privacy, security
- Language: Python
- Homepage: https://github.com/credona/age-decision
- Size: 113 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
- Roadmap: ROADMAP.md
Awesome Lists containing this project
README
Age Decision AntiSpoof
Age Decision AntiSpoof is the presentation attack detection service of the Age Decision ecosystem.
Responsibility
This repository owns still-image presentation attack detection and anti-spoof score production.
Scope
It estimates whether a face image appears to be a real human capture or a spoof attempt.
It does not expose raw model scores, heuristic details, or internal calibration details through the public API.
It does not perform identity verification, face recognition, document verification, age estimation, or active liveness challenge.
Version 2.3.0 tightens public contract governance: stable status endpoint coverage, standardized error responses, normalized request validation errors including missing_file for missing multipart uploads, and enforced privacy-first forbidden-field rules for anti-spoof outputs.
When to use this repository
- you need to detect spoof attempts (photo, screen, replay)
- you want to secure an age decision pipeline
When NOT to use this repository
- you need age estimation
- you need identity verification
- you need active liveness challenges
Documentation
- Repository: https://github.com/credona/age-decision-antispoof
- Usage: docs/usage.md
- Deprecation policy: docs/deprecation-policy.md
- Error model: docs/error-model.md
- Status contract: docs/status-contract.md
- Models and third-party notes: docs/models.md
- Benchmarks: docs/benchmarks.md
- Compatibility: docs/compatibility.md
- Security: SECURITY.md
- Global architecture and ownership: https://github.com/credona/age-decision/blob/main/docs/architecture.md
- Global scoring model: https://github.com/credona/age-decision/blob/main/docs/scoring.md
- Changelog: CHANGELOG.md
- Contributing: CONTRIBUTING.md
- Global project: https://github.com/credona/age-decision
Usage example
Run one anti-spoof decision request:
```bash
curl -X POST http://localhost:8001/check \
-H "X-Request-ID: test-request-001" \
-H "X-Correlation-ID: test-correlation-001" \
-F "file=@./test-face.jpg"
```
For setup, configuration, runtime options, Docker workflows, and full response details, see `docs/usage.md`.
License
This repository is released under the Apache License 2.0.
Third-party models may have their own upstream license and constraints.
See docs/models.md for details.