With the progress achieved to this date in mobile computing technologies, mobile devices are increasingly being used to store sensitive data and perform security-critical transactions and services. However, the protection available on these devices is still lagging behind. The primary and often only protection mechanism in these devices is authentication using a password or a PIN. Passwords are notoriously known to be a weak authentication mechanism, no matter how complex the underlying format is. Mobile authentication can be strengthened by extracting and analyzing keystroke dynamic biometric from supplied passwords. In this thesis, I identified gaps in the literature, and investigated new models and mechanisms to improve accuracy, usability and resilience against statistical forgeries for mobile keystroke dynamic biometric authentication. Accuracy is investigated through cost sensitive learning and sampling, and by comparing the strength of different classifiers. Usability is improved by introducing a new approach for typo handling in the authentication model. Resilience against statistical attacks is achieved by introducing a new multimodal approach combining fixed and variable keystroke dynamic biometric passwords, in which two different fusion models are studied. Experimental evaluation using several datasets, some publicly available and others collected locally, yielded encouraging performance results in terms of accuracy, usability, and resistance against statistical attacks. / Graduate / 2019-09-25

Indiana University-Purdue University Indianapolis (IUPUI) / Biometrics identi es/veri es a person using his/her physiological or behavioral characteristics. It is becoming an important ally for law enforcement and homeland security. Among all the biometric modalities, iris is tested to be the most accurate one. However, most existing methods are not designed for non-cooperative users and cannot work with o -angle or low quality iris images. In this thesis, we propose a robust multi-stage feature extraction and matching approach for non-cooperative iris recognition. We developed the SURF-like method to extract stable feature points, used Gabor Descriptor method for local feature description, and designed the multi- stage feature extraction and matching scheme to improve the recognition accuracy and speed. The related experimental results show that the proposed method is very promising. In addition, two template security enhanced schemes for the proposed non- cooperative iris recognition are introduced. The related experimental results show that these two schemes can e ectively realize cancelability of the enrolled biometric templates while at the same time achieving high accuracy.

The goal of this master thesis is to design and implement an approach for synthetic fingerprint generation from a biometric template. The thesis bases the solution on an existing fingerprint generator called SyFDaS developed at the Brno University of Technology, Faculty of Information Technology. Individual components of the generator had to be modified and automized to suit better the task of generating from a template. The end product enables the user to create a fingerprint without any intervention just by importing a template. The evaluation in this thesis presents results obtained by comparing the synthetic and original fingerprints using the VeriFinger algorithm. Entirely automatically created fingerprints achieved mixed results; however, manual adjustments of the parameters brought substantial improvements. Up to 72% of synthetic fingerprints reached the match by the VeriFinger. The results of the evaluation helped to identify weak points of the current solution. Based on these, the thesis proposes further steps to improve the success rate of automatic generation and the quality of other components.