CSE 891 Selected Topics: Biometrics, Spring 2005, MSU

Michigan State University
Spring 2005
CSE 891 - Selected Topics: Biometrics, 3 credits

Tu, Th: 3:00 pm - 4:20 pm, C207 WH

Instructor Information

Office: 3143 EB

Office Hours: M, W: 2 pm - 3 pm or by appointment

Phone: 355-9282

Email: jain@cse.msu.edu
TA: Xiaoguang Lu (lvxiaogu@cse.msu.edu, 3208 EB); Karthik Nandakumar (nandakum@cse.msu.edu, 3208 EB)
Office Hours: by appointment

Course Information

Homework 3 Assigned, Due April 14, 2005.

Hand-geometry system at Scott Air Force Base

Introduction
Prerequisites
Reference Books
Assigned Reading
Course Schedule
Homeworks
Case Study
Project
Grading
Exam Resources
Miscellaneous

Biometrics Research at MSU

Biometrics: Overview

Useful Biometrics Links

Introduction

Automatic and reliable identification of individuals for issuing official documents (e.g., passport and visa) and providing access to secure facilities (e.g., military base) and proprietary information (e.g., corporate websites) has become an essential part of our modern networked society. Biometric recognition systems utilize the physiological or behavioral characteristics of an individual for identification. By using biometrics, it is possible to establish an identity based on "who you are", rather than by "what you possess" (e.g., an ID card) or "what you remember" (e.g., a password). The events of 9/11 have generated huge interest in the design, deployment and evaluation of biometric systems. In this course we will study the design of various biometric systems based on fingerprints, voice, face, hand geometry, palmprint, iris, retina, and other modalities. Multimodal biometric systems that use two or more of the above characteristics will be discussed. Biometric system performance and issues related to the security and privacy aspects of these systems will also be addressed.

Prerequisites

Permission of the instructor. A background in probability and statistics, pattern recognition and image processing would be useful. However, students from non-computer science disciplines are encouraged to talk to the instructor about the course content and project.

Reference Books

D. Maltoni, D. Maio, A. K. Jain, and S. Prabhakar, Handbook of Fingerprint Recognition, Springer Verlag, 2003.
A. K. Jain, R. Bolle, S. Pankanti (Eds.), BIOMETRICS: Personal Identification in Networked Society, Kluwer Academic Publishers, 1999.
J. Wayman, A.K. Jain, D. Maltoni, and D. Maio (Eds.), Biometric Systems: Technology, Design and Performance Evaluation, Springer, 2004.
A number of books on biometrics have been put on the Assigned Reading in the Engineering Library. In addition, a number of journals, including Pattern Recognition, Pattern Recognition Letters, IEEE Trans. Pattern Analysis & Machine Intelligence (PAMI), IEEE Trans. Image Processing, IEEE Trans. Speech and Audio Processing and IEEE Security & Privacy Magazine routinely publish papers on biometrics.

Assigned Reading

Following books are on hold in the Engineering library for assigned reading for CSE 891.

S.Y. Kung, M.W. Mak, and S.H. Lin, Biometric Authentication: A Machine Learning Approach, Prentice Hall, 2005.

Ruud M. Bolle et al., Guide to Biometrics, Springer, 2004.

Paul Reid, Biometrics for Network Security, Prentice Hall PTR, 2004.

John D. Woodward et al., Biometrics, McGraw-Hill/Osborne, 2003.

John Chirillo and Scott Blaul, Implementing Biometric Security, Wiley, 2003.

Samir Nanavati, Michael Thieme, and Raj Nanavati, Biometrics: Identity Verification in a Networked World, John Wiley & Sons, 2002.

David Zhang (Ed.), Biometric Solutions for Authentication in an E-World, Kluwer Academic Publishers, 2002.

Anil K. Jain, Ruud Bolle, and Sharath Pankanti (Eds.), Biometrics: Personal Identification in Networked Society, Kluwer, 1999.

L.C. Jain et al. (Eds.), Intelligent Biometric Techniques in Fingerprint and Face Recognition, CRC Press, 1999.

Course Schedule

Mar 8, 10	SPRING BREAK
Jan 11	Overview of Biometrics: Definitions, biometric modalities, course outline A.K. Jain, A. Ross and S. Prabhakar, "An Introduction to Biometric Recognition" A. Jain, R. Bolle, and S. Pankanti: Introduction to biometrics Notes of Lectures 1-4 [in PDF, 6.5M]
Jan 13	Overview of Biometrics: Basic applications: access control, e-commerce, forensics...
Jan 18	Design of a Biometric System: Building blocks Form case study teams
Jan 20	Design of a Biometric System: Modes of operation HW1 Assigned Case study proposal due
Jan 25-27	Fingerprint A.K. Jain, L. Hong, S. Pankanti, and R. Bolle, "An identity authentication system using fingerprints" L. O'Gorman: Fingerprint verification Lecture Notes in PDF Minutiae Based Fingerprint Matching Non-minutiae Based Representations Fingerprint Enhancement Fingerprint Classification
Feb 1-3	Face (2D + 3D) J. Weng and D. L. Swets: Face Recognition Feb. 3: HW1 Due, HW 2 Assigned Lecture Notes [in PDF, 12M]
Feb 8	Lecture by Prof. Kumar
Feb 10	Lecture by Prof. Wayman
Feb 15	Iris J. Daugman: Recognizing Persons by Their Iris Patterns J. Daugman: How Iris Recognition Works Lecture Notes [in PDF, 4M]
Feb 17	Hand geometry R. Zunkel: Hand Geometry based Verification Lecture Notes [in PDF, 1M]
Feb 22	Midterm Exam
Feb 24	Signature V. Nalwa: Automatic On-line signature Verification Lecture Notes [in PDF, 1M]
Mar 1	Palmprint
Mar 3	Ear Biometrics M. Burge and W. Burger: Ear Biometrics Lecture Notes [in PDF, 1.8M]
Mar 15	Multimodal Biometrics: Limitations of unimodal systems, multibiometric scenarios, levels of fusion, system design A.K. Jain and A. Ross, "Multibiometric Systems"
Mar 17	Multimodal Biometrics: Score fusion techniques, score normalization, user-specific parameters, soft biometrics Lecture Notes [in PDF, 2M]
Mar 22	Case Study Presentations Biometrics in Banking Industry Biometrics in Computerized Patient Records
Mar 24	Case Study Presentations Biometrics in Credit Cards Biometrics in Mass Disaster Victim Identification
Mar 29	Voice J. Campbell: Speaker Recognition Lecture Notes [in PDF, 0.2M] HW 3 Assigned
Mar 31	Case Study Presentations Forensic Odontology DNA as a Biometric Identifier
Apr 5
Apr 7	Lecture by Dr. Pat Flynn
Apr 12	Case Study Presentations Evaluation of Face Recognition Systems Other biometric modalities: retina, gait, keystroke, teeth... R. Hill: Retina Identification M. S. Nixon, J. N. Carter, D. Cunado, P.S. Huang, and S. V. Stevenage: Automatic Gait Recognition M. S. Obaidat and B. Sadoun: Keystroke Dynamics Based Authentication Lecture Notes on Keystroke Dynamics
Apr 14	Biometric System Security: Attack points, solutions: watermarking, encryption, cancellable biometrics... J. Woodward: Biometrics: Identifying Law and Policy Concerns Lecture Notes HW 3 Due
Apr 19
Apr 21
Apr 26	Project Presentations
April 28	Final Project Report Due
May 2	Project Presentations

Homework

Homework 1: In this homework, you will study minutiae extraction, minutiae matching and the effect of fingerprint image quality on the matching performance.
Homework 2: In this homework, you will study a PCA based face recognition system. You will be provided with a public domain face database. Divide the database into training and test subsets.
1. Find the PCA representation of the face images in the training subset; also, show the basis images.
2. Calculate the scores for genuine and imposter matches in the test subset.
3. By changing the system decision threshold, generate a ROC curve; also, report the Genuine Accept Rates (GAR) at False Accept Rates (FAR) of 10%, 1%, 0.1%, 0.01%.
4. Report the equal error rate.
5. Show examples of faces that are correctly matched and incorrectly matched.
Homework 3: In this homework you will design a multimodal biometrics system, that will fuse face and fingerprint. You will use the fingerpirnt matcher of HW1 and face matcher of HW2 for 50 subjects, 10 images/subject (face & fingerprint)
1. Using the combinations of two score normalization and two fusion schemes, generate the genuine and imposter matching scores for a total of four multimodal systems; plot genuine and imposter distributions and ROC curves pertaining to multimodal as well as unimodal (face and hand geometry) systems.
2. Calculate the number of falsely rejected individuals if the decision threshold is set such that FAR=0.1% for multimodal and unimodal systems.
3. Elaborate on the performances of multimodal systems over unimodal systems. Which combination of normalization and fusion method leads to the highest performance? Explain.
4. Give examples of a user that is not correctly recognized using face or fingerprint alone but is recognized correctly using the combination of face and fingerprint.

Case Study

The students will work in teams of two on a particular deployment of biometric systems. Some suggested applications are:

Biometrics in Transportation (Airports, Railroads, etc)
1. http://www.biometricscatalog.org/asbwg/Files/Evaluating_Technology_Program_Plan.pdf
2. http://www.argus-solutions.com/pdfs/aviation_study.pdf
3. http://www.oki.com/en/otr/199/downloads/otr-199-R19-2.pdf
Biometrics in Commercial Applications (Grocery Stores, Hotels, etc)
1. http://www.kioskbusiness.com/JanFeb02/articles/dept3.html
2. http://asmallvictory.net/archives/001839.html
3. http://www.hotel-online.com/News/PressReleases2000_1st/Mar00_BiometricCurrency.html
4. http://www.findbiometrics.com/viewnews.php?id=603
5. http://www.hotelchatter.com/story/2004/10/12/191812/90/Title/Your_iris_is_the_key
Biometrics in Medical Services (Healthcare, Document Security, etc)
1. http://www.biometricsdirect.com/Documents/Case%20Study%20-%20NFMC.pdf
2. http://er1.org/docs/Informatics/Overview%20of%20Biometrics%20in%20Healthcare.pdf
Biometrics in Forensics (Criminal Justice, Disaster Victim Identification)
1. http://www.fbi.gov/hq/cjisd
2. http://www.interpol.int/Public/Forensic/
3. http://dgrc.org/dgo2004/disc/presentations/health/fahmy.pdf
Biometrics and Smart Cards (Federal Identify Cards, National ID, etc)
1. http://www.smartcardalliance.org/pdf/industry_info/smartcardhandbook.pdf (united states)
2. http://www.gemplus.com/pss/id_security/download/Oman_case_study_Aug04.pdf (oman)
3. http://www.identitycards.gov.uk (united kingdom)
4. http://www.asean-ssa.org/sssidsyssssphils.pdf (philippines)
Biometrics in Department of Defense
1. http://www.biometrics.dod.mil/
2. http://www.biometrics.dod.mil/SiteComponents/References.aspx
Standarization of Biometric Systems
1. http://www.bioapi.org
2. http://www.fmcsa.dot.gov/pdfs/tb99-003.pdf
3. http://www.frvt.org/DLs/FERET7.pdf
4. http://www.biometrics.dod.mil/documents/BRIEFS/Biometric-Standards.pdf
Students will make a presentation of the case study to the class in the middle of the semester.

Project

The students will work in teams of 2 on a class project. Possible topics for the project are:

Image quality assessment

Template characteristics with regard to the change over time

Performance comparisons between human and computer systems

Performance comparisons among different biometric sensors and modalities
Students will submit a project report at the end of the semester; further, they will present their project results to the class.

Grading

Course grade will be assigned based on scores of midterm exam, case study, homeworks, project, and class participation. Weights for these components are as follows: Midterm Exam (20%), Case Study (20%), Homework (30%), Project (30%). The cumulative score will be mapped to the course grade as follows: 90% or higher: 4.0; 85% to 90%: 3.5; 80% to 85%: 3.0 and so on.

The exam will be closed book. Makeup exams will be given ONLY if properly justified. Late homeworks and projects will not be accepted.

Please refer to MSU's policy on the Integrity of Scholarship. All homework solutions must reflect your own work. Failure to do so will result in a grade of 0 in the course.

Exam Resources

The midterm exam will test the understanding of basic biometric system concepts, such as false accept rate (FAR), fusion rules, template protection, fingerprint features, etc.

CSE 891 Home Page

Michigan State University Spring 2005 CSE 891 - Selected Topics: Biometrics, 3 credits