Biometric Quality Workshop
March 8-9, 2006
National Institute of Standards and Technology
March 17, 2006: Update
The workshop went ahead as planned. It was attended by 165 people, who saw the presentations of 41 speakers. The presentations are archived here. A summary will be released shortly.
Background
Performance of biometric systems is dependent on the quality of the acquired input samples. If quality can be improved, either by sensor design, by user interface design, or by standards compliance, better performance can be realized. For those aspects of quality that cannot be designed-in, an ability to analyze the quality of a live sample is needed. This is useful primarily in initiating the reacquisition from a user, but also for the real-time selection of the best sample, and the selective invocation of different processing methods. It is the key component in quality assurance management, and because quality algorithms often embed the same image (or signal) analyses needed to assess conformance to underlying data interchange standards, they can be used in automated image screening applications.
Quality analysis is a technical challenge because it is most helpful when the measures reflect the performance sensitivities of one or more target biometric matchers. NIST addressed this problem in August 2004 when it issued the NIST Fingerprint Image Quality algorithm, which was designed to be predictive of the performance of minutiae matchers. Since then NIST has been considering how quality measures should be evaluated, developing quality measures for other biometrics, and considering the wider use of such measures. In addition NIST is active in the new SC37 and M1 standardization activities on biometric quality and sample conformance.
Workshop Scope
The workshop is aimed at improving performance of biometric systems. It aims to assess current quality measurement capabilities and to identify technologies, factors, operational paradigms, and standards that can measurably improve quality. The workshop will be held over two days: Day One will consider applications, capabilities and standardization; Day two will consider research and development for all modalities, and evaluation.
- Day One: Applications
- How can Federal agencies and others benefit from quality assessment?
- What are the current and near-term applications?
- What are the use-cases, business-cases and economics?
- Are there applications beyond sample re-acquisition, quality assurance and survey, detection of evasion, and fusion?
- Day One: Capabilities
- What are the current and near-term capabilities?
- Can quality measurements adequately select the best sample from a stream?
- Is autocapture available or appropriate for all biometrics and applications?
- Can real-time measurements of quality be used to reduce FTE and FTA?
- What can be achieved by sensor design alone?
- To what extent do human factors drive quality?
- Can quality measurements flag an adverse or changed environment?
- Day One: Standardization
- What's the status of standardization? What else should be done?
- Does conformance to existing data format standards guarantee quality?
- Could Capture Best Practices annexes be added to, for example, ISO/IEC 19794-4?
- Would standard quality corpora be useful? Or standard reference algorithms?
- Is a standard interface needed?
- Day Two: Modalities
- What methods exist for assessment of face, finger, iris and speech quality?
- To what extent does good design guarantee quality?
- Can quality be related to liveness testing?
- Should quality just be a scalar snapshot summary of a sample?
- Does multimodal, multi-sample, or multi-sensor acquisition solve quality problems?
- Day Two: Evaluation
- How should quality measures be evaluated?
- Should quality be predictive of recognition performance, and evaluated as such?
- What are the relevant performance metrics?
- How should speed of operation be considered?
- Day Two: Research and Development
- What research is being done?
- Is the amount of research commensurate with its operational importance?
- Should research funding go into quality-by-design? Quality measurement? Or both?
Call for Participation
NIST solicits presentations, case studies, papers, and participation from any interested parties including vendors, systems architects, integerators, researchers, and users. Please contact the organizers with any proposed subject matter. We will then develop the program in response to the submissions received. We intend that material finally presented will be posted to the workshop web site. NIST encourages presentations and reports on preliminary work that participants plan to publish elsewhere.
We will produce a preliminary program early in the New Year and finalize it before March.
Who Should Attend
The workshop is aimed at individuals concerned with development, deployment and use of biometric quality infrastructure. Particularly the Day One program will be suited to representatives of Federal, State and local government agencies, policy makers from US and International organizations, technology providers, systems integrators and developers. The Day Two program will be directed at developers, integrators, engineers, researchers and students. We consider that the Day One content (i.e. needs and applications) should be of interest to those working in Day Two areas.