Overview
ISO/IEC TR 29189:2015 - Information technology - Biometrics - Evaluation of examiner assisted biometric applications defines the characteristics and performance-testing considerations unique to examiner‑assisted biometric systems. It explains how human examiners interact with automated matchers (for example in AFIS and latent fingerprint workflows), and why evaluation must account for both the automated algorithms and the examiner’s role in capture, encoding, matching adjudication and final decision.
Key Topics
- Scope and definitions: what constitutes an examiner and an examiner‑assisted application.
- Examiner‑assisted process model: stages where examiners typically intervene (capture, edit/encode, search parameters, result adjudication).
- System‑related factors: dependencies in the workflow, stage‑level vs end‑to‑end performance, measuring true operational performance, effects of prior probabilities and system confidence levels, and how automation affects human performance.
- Examiner‑related factors: examiner perception of system accuracy, usability and acceptance, training and expertise, workload, bias and individual differences.
- Performance evaluation methods: types of evaluation, relevant performance measures (accuracy, discrimination, bias, confidence, processing speed), and special considerations for scenarios like watch lists.
- Usability and reporting: qualitative observations, questionnaires, interviews, standardized reporting of results.
- Controls and challenges in testing: controls for examiner expertise and bias, test environment, variations in examiner input, and challenges when testing live operational systems or creating repeatable test conditions.
Applications
ISO/IEC TR 29189:2015 is practical for organizations that develop, deploy, evaluate or procure examiner‑assisted biometric systems:
- Forensic laboratories and law enforcement (AFIS, latent fingerprint comparison).
- Biometric system vendors and integrators designing examiner toolsets and human-in-the-loop workflows.
- Test and evaluation labs and conformity assessors creating performance test plans that combine human and automated components.
- Procurement officers and policymakers who need to specify performance criteria for examiner‑assisted deployments.
Who should use this standard
- Biometric evaluators, system architects, forensic examiners, usability specialists, and procurement teams who must assess or specify human-plus-system biometric performance rather than algorithm‑only metrics.
Related standards
- Related ISO/IEC biometric standards (for example, biometric data interchange and presentation‑attack detection standards) are complementary when defining data formats, security, and anti‑spoofing requirements for examiner‑assisted systems.
This report helps ensure evaluation of examiner‑assisted biometric applications reflects real operational performance by combining human factors, usability, and algorithmic metrics.