CS 9093 Biometrics

CS 9093 Biometrics – Course Syllabus

Course Description

Biometrics has emerged from relatively specialized use in the criminal forensics domain to more mainstream use for computer authentication, identification document security, and surveillance for public safety. This emergence has been accompanied by an expansion in biometric modality from mainly fingerprints to face, iris, hand, voice, and other novel biometrics.

This course concentrates on the unique advantages that biometrics brings to computer security, but also addresses challenging issues such as security strength, recognition rates, and privacy, as well as alternatives of passwords and smart cards. Students will gain knowledge in the building blocks of this field: image and signal processing, pattern recognition, security and privacy, and secure systems design. By the end of the course students will be able to evaluate and design security systems that include biometrics.

Course Objectives

In this course, students will learn to answer the following:

•What are the methods of biometrics?

•What are the devices of biometrics?

•How are these used for computer security?

•How do we design and build a secure system?

Three academic fields of concentration are taught in this course as related to biometrics: image processing, pattern recognition, and security and privacy. Students will learn how biometrics fits into these through lectures and assignments.

Successful completion of this course will prepare the student to do any of the following:

•Perform R&D on biometrics methods and systems

•Evaluate and design security systems incorporating biometrics

•Understand the technology of biometrics for public policy matters involving security and privacy.

Instructors

Lawrence O’Gorman, Ph.D.,
Nalini Ratha, Ph.D.,

/ Lawrence O'Gorman was at Bell Labs as a Distinguished Member of Technical Staff from 1984 to 1997. From 1997 to 2001, he was chief scientist for Veridicom, a maker of fingerprint capture chips, and is currently a Research Scientist at Avaya Labs in the area of multimedia security. He has published over 70 refereed papers, has 16 patents, and has been a contributor to 4 biometrics standards. He has served on US government scientific panels to NIST, NSF, and NAE, and to France’s INRIA. He is a Fellow of the IEEE and of the International Association for Pattern Recognition. He received the Ph.D. degree in Electrical Engineering from CarnegieMellonUniversity.
/ Nalini Ratha is a Research Staff Member at the IBMThomasJ.WatsonResearchCenter, Hawthorne where he leads the biometrics research activities. He has published more than 60 referenced journal and conference papers and is a co-inventor on 11 issued patents. He is an Associate Editor for Pattern Recognition journal and IEEE Trans. on Systems, Man and Cybernetics part B. He has co-chaired several biometrics conferences in recent years including IEEE AutoID, AVBPA, ICPR and BTAS. He is a Fellow of the IEEE (class of 2007). He graduated with a Ph. D. degree in computer science from MichiganStateUniversity.

Course Textbook

Guide to Biometrics (Springer Professional Computing)by R. Bolle, J. Connell, S. Pankanti, N. Ratha, Springer Press, 2003, ISBN0387400893

References (not required)

Biometrics:

Biometrics Personal Identification in Networked Society, Jain, Bolle, Pankanti (ed.s) 1999
Handbook of Fingerprint Recognition, Maltoni, Maio, Jain, Prabhakar, 2005
Automatic Fingerprint Recognition Systems, Ratha and Bolle (ed.s) 2003
Biometric Systems, Wayman, Jain, Maltoni and Maio (ed.s) 2004

Image Processing:

Practical Algorithms for Image Analysis: Description, Examples, and Code, Seul, O’Gorman, Sammon, 2000 (code from this will be available for Assignment 1)
Digital Image Processing, Gonzalez, Woods, 2002

Pattern Recognition:

Pattern Classification,Duda, Hart, Stork, 2000
Pattern Recognition, Theodoridis, Koutroumbas, 2006

Course Outline

Lecture 1 / Introduction to Biometrics
Lecture 2 / Everything about Fingerprints
Lecture 3 / Image and Signal Processing
Assignment 1 / Biometric Signal Processing
Lecture 4 / Other Biometric Modalities
Lecture 5 / Pattern Recognition
Lecture 6 / Recognition Measurement, Errors, and Statistics
Assignment 2 / Recognition measurement and testing
Lecture 7 / Identification System Errors and Performance Testing
Mid-Term Exam / Material to date
Lecture 8 / Multimodal Biometrics
NO CLASS / SPRING BREAK
Lecture 9 / Computer Security
Lecture 10 / Comparing Biometrics, Passwords, and Tokens
Assignment 3 / Secure biometric design project
Lecture 11 / Biometric Resources and Standards
Lecture 12 / Large Scale Biometrics and Systems Case Studies
Lecture 13 / Advanced Topics: Issues and Proposals
Lecture 14 / Secure Biometric Design Project Presentations
Final Exam / sometime during exam week

Course Policies and Procedures

There are three main components to this course:

  • Lecture – “Attend” the weekly on-line lecture, that is, open and go through the slides and listen to the audio if it is included. The Lectures will be made available to you on the e-College system by sometime Monday of the week of that lecture.
  • Homework – This will include assigned reading, questions to be solved, and discussion, and will accompany all (or practically all) lectures.The Homework will be made available to you at the same time as each Lecture. You will have until Tuesday (midnight) of the following week to submit the homework. Submit the Homework using the DropBox in the slot designated by the week it was assigned or add to the Discussion thread.
  • Assignments – These will be projects that are due 2-3 weeks after assigned (the date will be specified).Each is designed to put into practice and demonstrate knowledge and understanding of some of the main topics of the course. These are not designed to be overly time-consuming; they may require computer work, but will not require large-scale programming. Submit the Assignments using the DropBox in the slot designated by the Assignment number.

Late Policy. Homework is due by midnight on Tuesday following the week they are assigned. The assignments are due by midnight on the date specified (usually 2-3 weeks after assignment). Late homework and assignments will be penalized as follows: after midnight on the due day, the grade will drop by half (rounded down), and they will drop by a point on each late day subsequently until zero.

Grading Policies

Final Grades for this class will be based on homework, three assignments, a mid-term, and a final exam. Weightings will be applied as follows:

Homework / 15%
Assignment 1 (Image/Signal Processing project) / 15%
Assignment 2 (Pattern Recognition project) / 15%
Assignment 3 (Biometric System Design project) / 15%
Mid-Term Exam / 20%
Final Exam / 20%
Total / 100%

ePoly, Polytechnic UniversityCopyright 2007Page 1 of 4