Report

Teaser, summary, work performed and final results

Periodic Reporting for period 1 - AMBER (AMBER - enhAnced Mobile BiomEtRics)

Teaser

Biometric solutions (using modalities such as finger, face, iris, voice and signature) have seen increased prominence over the past decade with large-scale implementations. The use of biometrics on mobile devices is, however, at a critical point – end-users need to be...

Summary

Biometric solutions (using modalities such as finger, face, iris, voice and signature) have seen increased prominence over the past decade with large-scale implementations. The use of biometrics on mobile devices is, however, at a critical point – end-users need to be confident of biometric usage in terms of usability, privacy and performance; compromise in any one or more of these categories will lead to mistrust and a reluctance to adopt over-and-above conventional forms of authentication such as PIN or password. Although there are currently a number of EU research groups that are actively addressing challenges, AMBER will provide an original and innovative approach to the research topics.

Objectives:

• Address a range of current issues facing biometric solutions on mobile devices requiring research and development that fit into key areas of usability and reliability, novel interactions and privacy security and confidence.
• Within the ETN, collate Europe-wide complementary academic and industrial expertise to investigate these issues and provide a structure and environment to effectively facilitate training.
• Recognising the potential permanency of mobile biometric solutions, the ETN will train and equip the next generation of researchers
• Develop solutions and theory to ensure secure and efficient authentication whilst protecting privacy of citizens.

Work performed

Workpackage 1: Network management
- Recruited 10 early stage researchers
- Supported the mobility of the researchers to partner organisations
- Ensuring deliverable submission

Workpackage 2: Coordinated training
- Completed training program documentation
- Continued sharing of training materials
- Completed workshops and research conference

Workpackage 3: Dissemination
- Published 7 papers
- Members have presented at 7 events
- Twitter established

Workpackge 4: Mobile platform usability and reliability
- State of art on longitudinal studies deploying mobile devices related to age, psychological and usability aspects
- State of art on vulnerabilities in biometric systems with a focus on unsupervised environments
- State of art on touch screen interaction errors with mobile devices
- State of art on usability and accessibility of authentication processes
- Plan and setup of the longitudinal data collection

Workpackage 5: Novel solutions for mobile biometric interaction
- State of art on different multibiometrics architectures for mobile devices
- State of art on continuous authentication style
- Two academic papers published

Workpackage 6: Privacy, security and confidence in mobile biometric interaction
- State of art on nomadic use-case scenarios for mobile biometrics
- State of art on sensitive information flow in distributed mobile biometric scenarios and applications
- State of art on biometric template protection with emphasis on mobile devices
- State of art on vulnerabilities in biometric mobile devices deploying fingerprint, face and eye biometrics
- Four academic papers in review/completed

Workpackage 7:
- All ethics requirements completed
- On-going liaison with the Independent Ethics Advisor

Final results

\"ESR1 This project examines the use of swipe as a form of continuous authentication on mobile devices. The potential impact of this work is that developers will be able to understand the performance capabilities of mobile device swipe as a biometric when applied to different usage environments.

ESR2 This project explores the use of novel sensors (such as GPS, accelerometers and gyroscopes) embedded within mobile devices to establish how this can be used. The potential impact of this work will be to enable researchers and implementers to integrate background biometric assessment within mobile device operations.

ESR 3 Confidential transactions are increasingly performed on mobile devices, which calls for measures to prevent potential threats. However, as any technology is fraught with security and privacy issues, biometrics is no different. Having a proper biometric template protection scheme alleviates most of these security and privacy concerns. The final goal is the design of non-invertible biometric transforms based on Bloom filters operating in a multi-modal dataset within a mobile scenario.

ESR 4 Finger-vein-based biometric recognition has recently drawn the attention of many stakeholders. In this study, we take a step allowing the user to have the finger-vein patterns acquired while on the move. Preliminary results show that fusing hand vein data which were acquired simultaneously via multiple sensors would increase the accuracy of identification. The use of this device in an \"\"on-the-fly\"\" modality will increase the user convenience and its social acceptance.

ESR 5 Mobile biometrics consists in using biometric data in unsupervised environment, with the equipment (smartphones) of limited capabilities. Spoofing consists in substitution of real biometric data by the artificial ones. Potential impacts (including the socio-economic impact and the wider implications) of the project so far: The spoofing detection technology under investigation is based on ear photograph and touch ear images made with the use of smartphone.

ESR 6 Social problems and lifestyle habits can interfere with biometrics systems accuracy and challenge the reliability of biometrics recognition system. The research looks at equalizing the biometrics effects of such social factors, potential impact are to influence gender-specific facial expressions effects on face recognition systems, as well as facial recognition system’s reliability under simultaneous influence of mood variations, makeup and gender factors.

ESR 7 Several cases have been identified where biometric data captured by mobile devices discloses privacy-sensitive attributes. Potential impacts could be knowledge models on biometric data and their privacy aspects, future technologies for privacy-enabling technologies to protect biometric data processing of applications such as identification, and empirical studies on the impact of such models with regards to the authentication performance.

ESR 8 Individual’s privacy, especially children’s privacy, can be harassed by algorithms when information is extracting from biometric data such as voice, as mobile applications do with no authentication purpose. Our review on standards, security and privacy design aspects in biometrics showed that Transparency and Intervenability, are considered best practices, but optional. The plan is to provide a preliminary set of policies and guidelines in an industrial whitepaper for Informed Consent protocol design.

ESR 9 The most important expected result is to achieve methods for detecting presentation attacks, avoiding fraud in the use of biometric sensors. The impact of this work will guarantee the personal authentication using biometrics, and therefore increase citizen privacy.

ESR 10 The final target it to develop an evaluation methodology to assess on the usability and accessibility of authentication solutions. The expected results are to create a set of best practices for the design of authentica\"

Website & more info

More info: https://www.amber-biometrics.eu/.