Planning and Assessment of a Workshop on Undergraduate Education in Biometric Systems

Biometrics is the science of recognizing and authenticating people using their physiological features. The global biometrics market has a compound annual growth rate of 21.3 percent. There is much research interest in different biometric systems, which has led to increasing efforts in ensuring that biometrics is taught at the undergraduate level. The authors are in the final year of an NSF TUES Type 2 grant that is based on the theme of vertically integrating biometrics experiments throughout the undergraduate curriculum. Three universities have joined together in this effort. This paper describes the planning and assessment of a 3 day workshop that is based on the NSF funded effort. Fifteen faculty from across the country participated in this workshop. Undergraduate and graduate students also attended. The key points of the workshop included invited lectures and hands-on laboratory activities. The invited lectures included a tutorial on biometrics, detailed lectures on speaker recognition and a lecture on how to assess an educational intervention. The hands-on activities were presented such that the attending faculty could take them back to their respective universities. The workshop assessment results are very positive with respect to organization, quality of the invited lectures, quality of the handson activities and the social program. INTRODUCTION AND BACKGROUND Biometrics is the science of recognizing and authenticating people using their physiological features [1]. Border and immigration control, restricted access to facilities and information systems, cybersecurity, crime investigations and forensic analysis are just a few of the primary application areas of biometrics used by commercial, government and law enforcement agencies. The global biometrics market has a compound annual growth rate of 21.3 percent. There is much research interest in different biometric systems and this has led to increasing efforts in ensuring that biometrics is taught at the undergraduate level [2][3][4][5][6][7]. The authors are in the final year of an NSF TUES Type 2 grant that is based on the theme of vertically integrating biometrics experiments throughout the undergraduate curriculum [7]. Three universities have joined together in this effort [8][9][10][11][12]. Biometrics involves signal/image processing in a pattern recognition framework [13][14]. There are two types of biometric systems. Biometric identification (BID) systems identify a person among a set of people whereas biometric verification (BV) systems accept or reject a person’s claimed identity. Five biometric systems are considered in this project, namely, face, speech, iris, signature and fingerprint. Although fingerprint recognition forms the largest share of today’s market [15], there are practical tradeoffs with other systems as given in Table 1 [15]. This exemplifies the need for further research and educational activities pertaining to a variety of biometric systems. Face-based and speech-based recognition systems are particularly promising as their accuracy is improved. Table 1 Characteristics of Biometric Systems (taken from [15]) The desired characteristics of biometrics systems are [1]: • Permanence: The unique features of the biometric modality change very little over a person’s lifetime. System Accuracy Ease of Use User Acceptance Ease of Implementation Cost Fingerprint High Medium Low High Medium Face Low High High Medium Low Speech Medium High High High Low Iris Medium Medium Medium Medium High Signature Medium Medium High Low Medium • Unique: The probability of the physiological/behavioral characteristic of two different people being the same or resulting in the same biometric features is nearly impossible. • Easy integration: Biometric recognition technology can easily be integrated into existing security systems or operate as a standalone. No special purpose hardware is needed for integration to a personal computer. This is true of iris, speech and face based systems. • Cannot be spoofed: The biometric data are not susceptible to theft, loss or compromise. They cannot be artificially duplicated. • Non-Invasive and Quick: No invasive contact with a subject and quick to give a result. • Very high accuracy making it a method of choice for airport security and other biometrics applications. • Little memory required for data to be stored. • Reasonable cost. • High user acceptance. • Little performance degradation due to mismatched training and testing conditions. This is a very important research issue. As part of the NSF-funded project, the team conducted a 3-day workshop intended to disseminate information on the principles of biometrics and on strategies for introducing biometrics into the undergraduate curriculum. The remainder of this paper discusses the