- Course
- 2024-2025
- Semester
- 2
- ECTS
- 3
- Type
- Elective
- University
- UVigo
Subject objectives
This subject offers a general vision of the biometric identification techniques based on image and video. It delves into the most common ones: face, fingerprint and iris recognition.
The students will have comprised the common characteristics of the technicians of biometric identification, the evaluation metrics, the problems of practical implementation, the peculiarities of each biometric modality and the best way to combine them. Besides, they will have developed a critical analysis on the best working point for a concrete application, as well as an understanding of the peculiarities been due to demographic factors (sex, age, race, culture) in the design, development, evaluation and deployment of a solution of biometric identification.
Contents
-Basic principles of biometric identification
Identity versus biometric traits: Types of traits and biometric signatures.
Variance intra-class and *nter-class of the biometric signatures. Influence of the sensors in the different signatures.
Mathematical modelling of the biometric data: Extraction of characteristics. Compression. Representation versus Discrimination.
Recognition, Identification, Verification and Authentication. Types of errors: TER, ERR, FAR, FRR.
-Current biometric technologie
Physiological characteristics: fingerprints, iris, face, palm, retina, voice.
Behavioural characteristics: signature (static and dynamic), keystrokes.
Detection of alive sample.
Pros and conts in the use of each biométric trait.
-Facial recognition
Global technics (eigenfaces, fisherfaces) versus local technics (template matching, NCC, Elastic Bunch Graph Matching). The problem of the
variation of illumination and pose. The problem of the detection and normalisation.
Technicians of deep learning. Pros and cons.
-Fingerprint recognition
Representation of minucias. Hausdorff distance. Gabor. filters. Tolerance to deformations. Types of sensors.
-Iris recognition
Representation of the iris. Algorithm of Daugman. Algorithm of Wildes.
Recognition at a distance. Pros and cons of iris recognition.
-Multimodal recognition. Multibiometrics.
Combination of classifiers. Independent or correlated sources. Fusion of classifiers: intramodal, intermodal, algorithmic and scores-based. State of the art Systems using multimodal recognition and/or multibiometrics.
Basic and complementary bibliography
Basic Bibliography:
Big part of the material of study is based in scientific articles that will be left to student’s disposal in the LMS.
Complementary Bibliography:
Wayman, J.L., Jain, A.K., Maltoni, D., Maio, D. (Eds.), Biometric systems. Technology, Design and Performance Evaluation, 978-1-84628-064-1, 1, Springer, 2005.
Anil Jain, Ruud Bolle y Sarta Pankanti (Eds.), Biometrics. Personal Identification in Networked Society, 978-0-387-28539-9, 1, Kluwer Academic Publishers, 2006.
John Daugman, How iris recognition works, IEEE Transactions on Circuits and Systems for Vide, 2004
Competencies
CB8: Students are expected to be able to integrate knowledge and face the complexity of making judgments based on information that, being incomplete or limited, includes reflections on social and ethical responsibilities linked to the application of their knowledge and judgments.
CT1. Practice the profession with a clear awareness of its human, economic, legal and ethical dimension and with a clear commitment to quality and continuous improvement.
CT4 – Ability to understand the meaning and application of the gender perspective in different areas of knowledge and professional practice with the aim of achieving a more just and equal society.
CG4. Capacity for critical analysis and rigorous evaluation of technologies and methodology.
CG7. Autonomous learning ability for specialization in one or more fields of study.
CE2: Students are expected to know and apply machine learning and pattern recognition techniques applied to computer vision.
CE4. Conceive, develop and evaluate complex computer vision systems.
CE5: Students are expected to know how to analyze and apply state of the art methods in computer vision.
Teaching methodology
Laboratory practical:
Practices of the concepts showed in the masterclasses. They will be made with software accessible to all the students. Learning based in the resolution of practical cases and in small projects. The work will be in general autonomous and with independent study of the students. Some practices will be done in group and by means of cooperative learning. Intensive use of the virtual classroom
will be implemented.
During the face-to-face part of the practices of laboratory there will be a personalized attention to solve doubts and to help in the advances. During the asynchronous part an extensive use of the Learning Management Systems and the forums of debate will be implemented.
Lecturing:
Participatory master classes where the contents are exposed and the pros and cons that different options would have to solve practical cases will be advanced, leaving some open questions so that the students work them and arrive to their own conclusions.
During the master class the debate between the students will be forced and open questions will hang in the air.
Evaluation system
Objective questions exam (15):
Examination of short questions about the concepts studied and of individualized evaluation.
Evaluated Competences: CB3 CG4 CE2 CT1 CT4
Problem and/or exercise solving (15):
Examination of short problems on concepts and practices conducted and of individualized evaluation
Evaluated Competences: CB3 CG4 CE4
Laboratory practice (70):
Laboratory practices will have an evaluable part of individually or in groups depending on the type of practice.
Evaluated Competences:
CB3
CG4
CG7
CE2
CE4
Studying time and personal work
Recommended study time for students is about 2 hours per week. Additionally, we estimate that they should spend about 3 hours / week working in a number of assignments. All of these activities add up to around 60h/semester.
Subject study recommendations
Subjects that are recommended to be taken simultaneously
Advanced machine learming for computer vision/V05M185V01205
Advanced image processing and analysis/V05M185V01201
Subjects that it is recommended to have taken before
Image description and modeling/V05M185V01102
Fundamentals of machine leaming for computer vision/V05M185V01103
Fundamentals of image analysis and processing/V05M185V01101