The question of estimating uncertainty in measurement is fundamental to all scientific fields. In the field of automated human recognition, lack of repeatability and reproducibility of measurements has been noted since at least the 1970s. This study discusses current approaches to estimation of measurement uncertainty within the broader context of scientific philosophy and measurement science. The authors discuss the Duhem–Quine thesis on testing holism and international standards on estimating and reporting uncertainty in laboratory measurements, then apply these concepts to the estimation of uncertainty in technology, scenario and operational testing in biometrics. The authors advocate for moving beyond the calculation of ‘coverage’ intervals as defined in the ISO/IEC ‘guidelines for the expression of uncertainty in measurement’ to full application of the concepts of uncertainty assessment.

A two-stage geometric approach that is both scale and rotation invariant is implemented for extracting the unique features present in the surface of an ear image. As occlusion because of ear rings and hair significantly affect the efficiency of ear recognition process, only the middle portion of the ear is considered in this work. The resultant matching scores are compared against a threshold to make a decision for authenticating a person. It is found that the fused scores obtained from the two levels of feature extraction enhance the recognition accuracy compared with that of the individual stages. Finally, particle swarm optimisation technique is applied on the matching scores in order to optimise the fusion parameters such as decision threshold and weights. It results in further improved verification rates compared with the fusion of scores without optimisation. Thus, the proposed method works on partial ear images and demonstrate the presence of more unique features in the middle part of the ear (as seen by the increase in recognition accuracy) and the method also aids in reducing the computation time.

Tactical biometric devices are used to establish the identity of individuals of interest in various military and law-enforcement scenarios. Most testing of these devices has been conducted in laboratory settings rather than in operationally-realistic tactical scenarios. This study describes an experiment which can viably replace this paradigm by measuring the performance of handheld biometric devices in a variety of tactical environments. The experimental procedure assessed the collectability, quality and matchability of images collected in operationally-realistic scenarios. Iris recognition accuracy was measured using several commercial algorithms. Results illustrate performance degradation in operational results relative to laboratory results; the collection limitations of the devices in operationally-realistic settings; and the effects of operators, subjects, devices and environments on performance. The authors believe that this experiment is unique in its exploration of these elements and that the powerful results presented suggest a need for refinement of design and procurement criteria.

This study examines session variability modelling for face authentication using Gaussian mixture models. Session variability modelling aims to explicitly model and suppress detrimental within-class (inter-session) variation. The authors examine two techniques to do this, inter-session variability modelling (ISV) and joint factor analysis (JFA), which were initially developed for speaker authentication. We present a self-contained description of these two techniques and demonstrate that they can be successfully applied to face authentication. In particular, they show that using ISV leads to significant error rate reductions of, on average, 26% on the challenging and publicly available databases SCface, BANCA, MOBIO and multi-PIE. Finally, the authors show that a limitation of both ISV and JFA for face authentication is that the session variability model captures and suppresses a significant portion of between-class variation.

Biometric template protection is suitable with the widespread deployment of biometric authentication schemes. Template protection methods are used to ensure the diversity and the security of biometric data, by avoiding storage and misuse of the original template. This study provides an evaluation of the accuracy performance of biometric template protection methods, by revisiting experiments on the biohashing algorithm on fingerprints. It is shown how and why experimental results can be completely falsified with only five random bits.