Files
boc/quixzoom-capture-pipeline/training/pipeline-report.md
T
Bernt bae705aa97 ARCHITECTURE: NFC roadmap, edge AI, audit logging
- Add NFC ePassport roadmap (ICAO 9303, eIDAS)
- Add TensorFlow.js edge face detection (BlazeFace)
- Add structured audit logger (GDPR-compliant)
- Risk scoring support

Part of KYC Apple Native UX v1.1.0
2026-06-29 16:24:48 +00:00

803 B

QUIXZOOM Active Learning Report

Generated: 2026-06-28T12:40:14.401386

Metrics

Metric Value Target
Raw images 0 100,000
Weak annotations 0 1,000,000
Human reviewed 0 100,000
Gold annotations 0 100,000
Training iterations 0 -
mAP@50 0.00 0.90
mAP@50-95 0.00 0.70

Pipeline Status

  • Raw data collection: NEEDS DATA
  • Weak supervision: NEEDS DATA
  • Human review: NEEDS REVIEW
  • Gold dataset: NEEDS MORE
  • Model training: NOT STARTED
  • Deployment: NEEDS IMPROVEMENT

Next Steps

  1. Collect more raw data from Bangkok and Torrevieja
  2. Run weak supervision on new data
  3. Prioritize human review of uncertain predictions
  4. Train model when gold dataset reaches 1000
  5. Deploy when mAP@50 reaches 0.9