Signature Biometric Systems: Usability Assessment

Biometric recognition for human identification and verification (using modalities such as face, iris and fingerprint) is an increasingly attractive and practically viable way to improve system security for a range of application domains including travel and national identification, physical access security and on-line management. The human automatic signature verification (ASV) modality is a key player in the biometrics market with particularly strong uptake in financial, legal and retail security. While considerable research has been conducted into the assessment of the algorithmic performances of biometric technologies, a range of new issues concerning the measurement of usability factors within biometric systems have emerged which require urgent attention following large scale deployment of public systems. A vital component in selecting and implementing a biometric system is to ensure that it performs adequately to meet the security requirements of the application. Conventionally this has been measured using a number of error rate metrics including false accept rate (FAR - the number of impostors erroneously verified as genuine by the system) and false reject rate (FRR - the number of genuine subjects rejected as being impostors), alongside other metrics assessing the rate of successful enrolment and biometric sample detection by a sensor. Although widely used to provide an indication of computer-based system performance, these statistics do not quantify other errors contributing to the overall performance of the system, specifically i) relating to the environment within which the system is used and ii) how the test subject interacts with the system. As biometrics become more widespread and the community becomes more experienced in testing and evaluation, standardised testing methodologies and their associated evaluation metrics need to be re-examined. This application seeks funding for the development of research collaboration links between the Image and Information Engineering Research Group at the University of Kent and the Biometrics Standards, Performance and Assurance (BSPA) Laboratory at Purdue University, US. Both partners in this proposal have recently independently worked on novel methods for the performance evaluation of biometric systems and, recognising this complementarity in methods, have held informal research planning discussions electronically as part of the ISO International Standardisation processes within which both centres are actively involved.In this study we will combine the research strengths at the two institutions to assess and apply the Human Biometric Sensor Interaction (HBSI) framework to the ASV modality by establishing how users interact with signature devices and systems beyond the normal system performance metrics.Specifically this application is to enable the PI to travel to Purdue to: 1) facilitate further interaction between the two research centres, 2) to define an experimental procedure to investigate the usability of ASV systems using the HBSI model, 3) to implement a cross-Atlantic data collection exercise to enable ASV system interaction to be analysed and 4) to facilitate the presentation of talks, teaching material and generate future research initiatives.

The biometrics (automatic identification/verification by physical and behavioural traits) industry continues to expand with sustained growth forecast to reach $11 billion worldwide by 2017 (representing a market tripling between 2010 and 2017), indicating increased reliance on systems for a wide range of applications including travel documentation, personal and property security and identification and surveillance and monitoring. This proposal focuses on the automatic signature verification (ASV) modality, the use of which is expected to double in the period to 2017, primarily in the financial and legal domains. The UK has and continues to be amongst the world leaders in the development of ASV systems, most recently leading developments in ISO standardisation. This Overseas Travel Application aims will robustly assess the human-usage aspects of ASV systems, building directly upon complementary skills of the two partner centres. Biometric system usage issues have been identified as a core area of research within the community requiring studies to assess how a user of a system interacts with a technology, enabling a deeper understanding of factors restricting performance and experience of use relating to device interaction, user-interface design, speed of use, sampling errors, user expectations, etc. This work is of central importance given that commercial implementation are being used by the general public on a daily basis, often with limited user training and high expectations. The findings of this study will potentially provide answers to the following stakeholder groups: 1. Biometrics and security industry: the UK is a major international player in biometric security systems development, integration and implementation. Most projects that are undertaken by UK industry and academia are global systems, therefore potential impact is high. This group will benefit from a deeper understanding of the interaction design of ASV systems, how a user donates a signature on a range of devices, HCI considerations and interaction speed. These issues will be considered in a quantifiable form and will serve as best practice for implementation. 2. System commissioners and end-users: both public and private agencies are increasingly turning to biometric systems to provide front-line security solutions with the result that the pool of consumers is growing exponentially. All public systems must be intuitive in terms of usability and must also give the user a high level of confidence in the process, otherwise likely re-use is significantly reduced. As this work will directly address issues of system interaction across a wide community user-base, this will be of direct interest to commissioners of signatures systems. Across both of these stakeholder groups the work also addresses an aim of the UK Digital Economy Programme to 'Support and enhance a UK academic community of Digital Economy researchers, who understand the technology and how people use it and what the impact is.' The applicants are very strongly positioned to ensure that the impact of the work is maximised. The applicants are respected and established within the academic community (to enable dissemination through academic channels) but, more importantly in terms of impact, are directly involved as committee chairs within their national bodies on biometrics standardisation (UK BSI IST/44 and US INCITS/M1). These bodies bring together government, industry, end-users and academic representatives and will enable propagation of results and recommendations to both a national and international audience. Clearly, therefore, combining the mutual expertise of the two research groups, the recognised urgent requirement of the research areas within the academic and commercial communities and the commercial/governmental connections of the investigators, increases the possibility of high impact of the work.