Not started
Work Package
WP7 — Move through document verification and biometric verification functions and their aggregation (Lead: THALES)
Goals & Objectives
- Enhance system security through advanced biometric verification and anti-spoofing measures.
- Integrate ID document verification capabilities to support comprehensive identity authentication.
- Ensure compliance with GDPR and privacy standards throughout the biometric verification process.
Tasks
- Task 7.1: Biometric face (multipart-based) and fingerprint recognition (THALCZ (L), UL; M7-30) Multipart-based face recognition divides facial images into segments (e.g., face, eyes, ears) to analyse them independently, improving accuracy and resilience against occlusions or spoofing. Deep models combine these parts into a robust facial profile while ensuring GDPR-compliant data encryption and privacy. Fingerprint recognition leverages smartphone sensors and native APIs (e.g., Touch ID) to authenticate users securely. Data processing occurs locally within secure hardware, ensuring privacy, reducing costs, and enhancing user convenience.
- Task 7.2: Face data injection, presentation attack analysis, Liveness detection smartphone face and prevention (CLR (L), UL, THALES, THALCZ, AIT, IDENTY, VSENS; M7-30) Face data injection tests system integrity, while presentation attack analysis detects spoofing attempts like masks or photos. Liveness detection on smartphones ensures genuine interactions by analysing micro-movements, texture, or depth. Prevention relies on advanced algorithms, multi-modal biometrics, and secure, GDPR-compliant data handling.
- Task 7.3: Passengers ID document and breeder document verification capability made available in extended digital wallet though NUI node emulation and ESS node emulation with VIS visa approval emulation and ETIAS (non visa required) emulation (THALES (L), UL, XLAB, IDENTY; M7-30) ID document verification at standstill involves using specialized tools to authenticate foundational documents, like birth certificates and IDs, which establish an individual’s identity. At a stationary checkpoint or kiosk, the verification system analyses these documents’ physical and digital security features such as watermarks to ensure authenticity. Advanced OCR and biometric cross-checks can also be performed to validate consistency across related documents, minimizing the risk of fraud. The standstill setup allows for detailed examination, providing reliable, accurate verification that aligns with compliance and security requirements while maintaining user privacy and data protection.
- Task 7.4: Co-time stamping and Co-localization analysis (FHG (L), CAFA, VSENS, IDEMIA; M7-30) Co-time stamping and co-localization analysis involves correlating the time and location data of multiple entities or devices to verify their presence together at a specific moment and place. FHG, CAFA, VSENS, CAFA and IDEMIA apply it in all modalities to localization and times for ultrasound smartphone localization, heartbeat person count, travellers’ selfie and fingerprint images, and external multispectral camera-based person face verification, tracking and best image extraction. This technique is essential for applications requiring strict synchronization and spatial verification, such as verifying group and person movement patterns, validating proximity in security contexts, or confirming simultaneous presence for event tracking. By analysing time stamps and geolocation data across sources, the OnMoveID system can determine the alignment and authenticity of recorded interactions, identify discrepancies or confirm credible and legitimate co-presence. Co-time stamping ensures that the timing is precise, while co-localization confirms spatial accuracy, creating a robust framework for analysing coordinated activities while supporting data integrity and compliance with privacy standards, including that OnMoveID verification is limited to walk-through, drive-through and short stop lanes only in space and time.
- Task 7.5: Data management plan & Pre-Registration and border crossing acceptance assessment (risk and rule based) (FHG (L), UL, THALES, THALCZ, XLAB, CAFA, IDENTY, VSENS, IDEMIA; M7-34) FHG develops together with the task partners a function that allocates information gathered in each border crossing step (at-home/on-move registration, back-up registration at border) and during move-through at border (walk-through, drive-through, or short bus stop) sufficient for decision making of acceptance or refusal. Minimum and sufficient information is collected in privacy-preserving way supported by the respective functional partners (THALES/CZ, FHG, CAFA, IDEMIA and VSENS) compliant with credentials as provided within the digital EUID digital wallet framework supported by IDENTY and XLAB. For instance, for registration: digital wallet securely operational, identify verified of digital identity, of physical passport, and of breeder document, and biometric face and fingerprint of identity documents fit to selfie and fingerprints taken by passenger for registration. Using a similar approach, the partners will build on results of T7.5 to generate further independent credentials for all steps and modalities. The decision making will be rule and risk-based according to Border Guards’ requirements as identified in WP3 and further develop approaches as developed within the EU projects XP-DITE and TRESSPASS. Decisions will be transparent and communicated privacy preserving and securely to travels within extended digital wallet app. Border Guards are informed via their extended digital wallet app on various devices. At border the decision covers acceptance of identity/entry refusal. Border Guards are informed regarding issues that could be resolved at site, e.g. to take an external face image for additional checking or to conduct human person verification. The acceptance assessments also consider information as obtained from legacy and emerging systems interfacing with OnMoveID. To this end, acceptance recommendations/decisions/issues of a representative list of such systems as identified in WP3 is used as provided, e.g. EES, Visa System, mock-up searched person lists. The function will flag to Border Authorities those needing additional human screening while automating and expediting low-risk entries that obey all rules.