Class activities will be recorded here.

Complete lecture notes and any supplemental material will be made available immediately following each class.

Week 1 : August 26 - August 28
  • Wed.
    Introduction and Biometric Basics 1
    Welcome to Biometrics! Our first lecture covers:
    • Course overview and expectations
    • Biometrics basics
    Fri.
    Biometrics Basics 2
    The second lecture covers case studies to put biometrics research into context:
    • A brief history of the last decade in biometrics applications
    • The ethics of biometrics
    • Practical security concerns
Week 2 : August 31 - September 4
  • Mon.
    Biometrics Basics 3
    The third lecture in this unit introduces definitions common to the discipline of biometrics:
    • Data source assumptions
    • Questionable claims of uniqueness and examiner infallibility
    • Matching modes
    • A formal model of recognition
    Wed.
    Biometrics Basics 4
    The fourth lecture in this unit finished introducing definitions and began to describe error statistics:
    • A formal model of recognition
    • The enrollment process
    • Score distributions
    Fri.
    Biometrics Basics 5
    The fifth lecture in this unit describes error statistics:
    • Security problems related to false matches
    • Performance curves
    • Consideration for machine learning
Week 3 : September 7 - September 11
  • Mon.
    Biometrics Basics 6 and Face Recognition 1
    This lecture wrapped up the biometrics basics unit and kicked off the face recognition unit:
    • Information retrieval statistics
    • The history of face recognition
    • Benchmark data sets for face recognition evaluation
    Wed.
    Face Recognition 2
    The second lecture in this unit introduced face detection:
    • Haar-like Features
    • Boosting
    • Attentional Cascades
    Fri.
    Face Recognition 3
    The third lecture in this unit described a recent approach for facial landmark detection:
    • Cascaded Regression
Week 4 : September 14 - September 18
  • Mon.
    Face Recognition 4
    The fourth lecture in this unit focused on problems in unconstrained face recognition:
    • Discussion of Homework #1
    • Contemporary data sets for face recognition evaluation
    Wed.
    Face Recognition 5
    The fifth lecture in this unit continued to develop the face pre-processing pipeline:
    • Face detection that is tolerant to pose, scale, and occlusion
    • Face alignment
    Fri.
    Face Recognition 6
    The sixth lecture in this unit continued to develop the face pre-processing pipeline:
    • Face alignment (in-plane rotation)
    • Face alignment (out-of-plan rotation, analysis-by-synthesis)
    • Lighting normalization
Week 5 : September 21 - September 25
  • Mon.
    Face Recognition 7
    The seventh lecture in this unit finished describing the face pre-processing pipeline and began to introduce features for face recognition:
    • SQI lighting normalization
    • Eigenfaces
    • Fisherfaces
    Wed.
    Face Recognition 8
    The eighth lecture in this unit introduced hand-tuned features for face recognition:
    • Gabor Filters
    • V1-Like Model
    • EBGM
    • SIFT
    Fri.
    Face Recognition 9
    The ninth lecture in this unit finished our survey of hand-tuned features for face recognition:
    • SIFT
    • HOG
    • LBP
    • Fisher Vectors
Week 6 : September 28 - October 2
  • Mon.
    Face Recognition 10
    The tenth lecture in this unit introduced deep learning for face recognition:
    • Neural networks basics
    • Convolutional neural networks
    Wed.
    Face Recognition 11
    The eleventh lecture in this unit described deep learning for face recognition:
    • Parts of a convolutional neural network
    • Google FaceNet
    Fri.
    Face Recognition 12
    The twelfth lecture in this unit covered different imaging modalities:
    • 3D Face Recognition
    • Considerations for surveillance
Week 7 : October 5 - October 9
  • Mon.
    Face Recognition 13
    The final lecture in this unit covered visual attributes and video:
    • Describable visual facial attributes
    • Attribute-based face search
    • Face recognition in video
    Wed.
    Fingerprint Recognition 1
    This first lecture in unit 3 introduced us to fingerprint recognition:
    • The history of fingerprint recognition
    • Features on the finger
    Fri.
    Fingerprint Recognition 2
    This second lecture in unit 3 discussed fingerprint uniqueness and features:
    • Where did Galton's error estimate come from?
    • Detecting singular points
    • Fingerprint classification
    • Fingerptint minutiae
Week 8 : October 12 - October 16
  • Mon.
    Fingerprint Recognition 3
    The third lecture in unit 3 discussed fingerprint acquisition and storage:
    • Fingerprint sensors
    • Compression
    • Storage standards
    • Benchmark data sets
    Wed.
    Fingerprint Recognition 4
    The fourth lecture in unit 3 discussed fingerprint pre-processing:
    • Benchmark data sets
    • Distortions
    • Segmentation
    • Matching at various levels
    Fri.
    Fingerprint Recognition 5
    The fifth lecture in unit 3 discussed fingerprint matching:
    • Detecting minutiae
    • Matching minutiae
    • Longitudinal studies
Week 10 : October 26 - October 30
  • Mon.
    Iris Recognition 1
    The first lecture in unit 4 introduced the concept of iris recognition:
    • Anatomy and physiology of the iris
    • Overview of iris recognition process
    • History of iris biometrics
    Wed.
    Iris Recognition 2
    The second lecture in unit 4 wrapped up the history of iris biometrics and introduced pre-processing:
    • History of iris biometrics
    • Acquisition
    • Segmentation
    Fri.
    Iris Recognition 3
    The third lecture in unit 4 explained the iris recognition process:
    • Iris unwrapping
    • Masking
    • Iris codes
    • Security analysis
Week 11 : November 2 - November 6
  • Mon.
    Iris Recognition 4
    The fourth lecture in unit 4 looked at problematic conditions for iris recognition:
    • Pupil dilation ratio
    • Contact lenses
    Wed.
    Iris Recognition 5
    The fifth lecture in unit 4 continued to look at problematic conditions for iris recognition:
    • Iris template aging
    Fri.
    Multi-Biometrics 1
    The first lecture in unit 5 introduced the area of multi-biometric fusion:
    • Problems with individual modalities
    • Multi-biometric systems
    • Levels of fusion
Week 12 : November 9 - November 13
  • Mon.
    Multi-Biometrics 2
    The second lecture in unit 5 detailed the different levels of multi-biometric fusion:
    • Feature-level fusion
    • Rank-level fusion
    • Decision-level fusion
    Wed.
    Multi-Biometrics 3
    The third lecture in unit 5 examined decision- and score-level fusion methods:
    • Dempster-Shafer Theory
    • Score-level fusion techniques
    Fri.
    Multi-Biometrics 4 and Biometric Security 1
    This lecture wrapped up unit 5 and introduced biometric security:
    • Performance of score-level normalizations
    • An overview of biometric system security
Week 13 : November 16 - November 20
Week 14 : November 23 - November 27
  • Mon.
    Spoofing 2
    This lecture described countermeasures to biometric spoofing:
    • Subversive vs. suspicious action
    • Preventing spoofing attacks against fingerprint recognition
    • Preventing spoofing attacks against iris recognition
Week 15 : November 30 - December 5
  • Mon.
    Template Protection 1
    The first lecture in unit 8 examined the interplay between security and privacy:
    • Is there a tradeoff between security and privacy?
    • Nymity, linkability, observability
    Wed.
    Template Protection 2
    The second lecture in unit 8 continued to set the stage for template protection algorithms:
    • Requirements for template protection
    • Akerlof's Lemons market and its relationship to biometric security
    Fri.
    Template Protection 3
    The third lecture in unit 8 described template protection algorithms:
    • Types of template protection algorithms
    • Attacks against template protection algorithms
    • Fuzzy vaults
Week 16 : December 7 - December 9
  • Mon.
    Template Protection 4
    The fourth lecture in unit 8 described template protection algorithms:
    • Fuzzy commitment
    • Fuzzy extractors
    • Recovable biotokens
    Wed.
    Template Protection 5
    The final lecture in unit 8 described infrastructure for protocols incorporating template protection:
    • Biometrics in the cloud
    • Biometric key infrastructure