IA/0131/EN - BIOMETRIC TECHNOLOGIES AND BEHAVIOURAL SECURITY
Academic Year 2021/2022
Free text for the University
GIAN LUCA MARCIALIS (Tit.)
- Teaching style
- Lingua Insegnamento
|[70/90] COMPUTER ENGINEERING, CYBERSECURITY AND ARTIFICIAL INTELLIGENCE||[90/00 - Ord. 2018] PERCORSO COMUNE||5||50|
The main objective of the course is to give the basic concepts and algorithms on biometric technologies, with reference to the advanced videosurveillance and behavioural biometrics.
Knowledge and understanding
The student will know the basics of biometrics as a science of univocally identifying people by physiological or behavioural characteristics. The applications of this science will involve the cybersecurity and artificial intelligence domains.
Ability to apply knowledge and understanding
The student will be able to understand the organization and the working logic of any biometric system, to develop algorithms for the physiological and behavioural personal recognition from images and videos. The learnt principles will be applied by Matlab, Phyton and C programming languages.
The student will be able to assess both the adequacy of biometric systems and the data structures and approaches to personal recognition problems and of physical and logical security.
The student will be able to converse with experts on cybersecurity, artificial intelligence, videosurveillance, and describing the solution process for physical and logical security problems.
Ability to learn independently
The student will be able to learn advanced methods and new concepts for the design and analysis of biometric technologies autonomously.
Knowledge of language programming with preference on Matlab, Python and C. Basics of machine learning and artificial intelligence.
Introduction to biometric technologies. Modules of a biometric system.
Review of machine learning, pattern recognition and artificial intelligence.
Fingerprints. Modules. Features. Classification, identification, authentication. Presentation attacks detection. Data sets.
Faces. Modules. Features. Classification, identification, authentication. Facial expressions. Presentation attacks detection. Data sets.
Multimodal biometrics. Taxonomy. Fusion approaches. Data sets.
Other biometric technologies overview: EEG, gait, iris, retinal scan, palmvein and palmprint.
Advanced systems for biometrics: deep learning approaches. State of the art. Methods and algorithms. Learning and auto-encoding.
Crowd analysis and scene interpretation. Flows analysis. Design of an advanced videosurveillance system. Applications.
Frontal lectures: 29 hours.
Laboratory: 21 hours.
To meet specific educational needs related to the epidemiological situation, the possibility of live streaming lessons or recordings of the same available online is provided.
Furthermore, the exercises can be carried out through forms of remote interaction with the available IT supports.
Verification of learning
The exam consists of an assigned project whose delivery is according to a fixed deadline.
The project topic concerns one of the topics dealt with in the course and is aimed at understanding if the candidate has the theoretical knowledge taught during the lessons and the practical skills necessary to face the laboratory exercises during the course. The project can be realized in any programming language (C, Python, C #, Matlab recommended).
The student's ability to implement the design modules of biometric systems or video surveillance systems are assessed according to the quality of the product code and the ability to carry out the project according to the constraints required in terms of performance parameters seen in class.
Once the project has been completed, the candidate is invited to illustrate the characteristics of his solution in an oral interview.
COVID-19 emergency. The project is assigned as an activity through the MS Teams platform. The student receives communication via email issued on esse3 regarding the link to access the test. The rules were however indicated on the course website:
11) A. Jain et al., Handbook of Biometrics, Springer, https://www.springer.com/gp/book/9780387710402
2) B. Bhanu and A. Kumar, Deep learning in biometrics, Springer, https://www.springer.com/gp/book/9783319616568
3) K. Saeed, New direction in behavioural biometrics, CRC Press, https://www.crcpress.com/New-Directions-in-Behavioral-Biometrics/Saeed/p/book/9781498784627
4) V. Murino et al., Group and crowd behavior for computer vision, Academic Press, https://www.sciencedirect.com/book/9780128092767/group-and-crowd-behavior-for-computer-vision#book-info
5) D. Maltoni et al., Handbook of fingerprint recognition, Springer, https://www.springer.com/gp/book/9781848822535
6) H. Liu, Face Detection and Recognition on Mobile Devices, Elsevier, https://www.elsevier.com/books/face-detection-and-recognition-on-mobile-devices/liu/978-0-12-417045-2
7) M. Vatsa et al., Deep learning in biometrics, CRC Press, https://www.crcpress.com/Deep-Learning-in-Biometrics/Vatsa-Singh-Majumdar/p/book/9781138578234
Slides of the course to support teaching and learning will be delivered.
Other information about the course at the people page: