You are here
Search results
(1 - 2 of 2)
- Title
- Semi-Adversarial Networks for Imparting Demographic Privacy to Face Images
- Creator
- Mirjalili, Vahid
- Date
- 2020
- Collection
- Electronic Theses & Dissertations
- Description
-
Face recognition systems are being widely used in a number of applications ranging from user authentication in hand-held devices to identifying people of interest from surveillance videos. In several such applications, face images are stored in a central database. In such cases, it is necessary to ensure that the stored face images are used for the stated purpose and not for any other purposes. For example, advanced machine learning methods can be used to automatically extract age, gender,...
Show moreFace recognition systems are being widely used in a number of applications ranging from user authentication in hand-held devices to identifying people of interest from surveillance videos. In several such applications, face images are stored in a central database. In such cases, it is necessary to ensure that the stored face images are used for the stated purpose and not for any other purposes. For example, advanced machine learning methods can be used to automatically extract age, gender, race and so on from the stored face images. These cues are often referred to as demographic attributes. When such attributes are extracted without the consent of individuals, it can lead to potential violation of privacy. Indeed, the European Union's General Data Protection and Regulation (GDPR) requires the primary purpose of data collection to be declared to individuals prior to data collection. GDPR strictly prohibits the use of this data for any purpose beyond what was stated. In this thesis, we consider this type of regulation and develop methods for enhancing the privacy accorded to face images with respect to the automatic extraction of demogrpahic attributes. In particular, we design algorithms that modify input face images such that certain specified demogrpahic attributes cannot be reliably extracted from them. At the same time, the biometric utility of the images is retained, i.e., the modified face images can still be used for matching purposes. The primary objective of this research is not necessarily to fool human observers, but rather to prevent machine learning methods from automatically extracting such information. The following are the contributions of this thesis. First, we design a convolutional autoencoder known as a semi-adversarial neural network, or SAN, that perturbs input face images such that they are adversarial with respect to an attribute classifier (e.g., gender classifier) while still retaining their utility with respect to a face matcher. Second, we develop techniques to ensure that the adversarial outputs produced by the SAN are generalizable across multiple attribute classifiers, including those that may not have been used during the training phase. Third, we extend the SAN architecture and develop a neural network known as PrivacyNet, that can be used for imparting multi-attribute privacy to face images. Fourth, we conduct extensive experimental analysis using several face image datasets to evaluate the performance of the proposed methods as well as visualize the perturbations induced by the methods. Results suggest the benefits of using semi-adversarial networks to impart privacy to face images while still retaining the biometric utility of the ensuing face images.
Show less
- Title
- Development of novel computational techniques for the study of biomolecular systems using molecular dynamics simulation
- Creator
- Mirjalili, Vahid
- Date
- 2014
- Collection
- Electronic Theses & Dissertations
- Description
-
In this dissertation, we have developed novel computational techniques that have been effectively utilized to extend our knowledge of proteins and lipid membrane systems. Application of molecular dynamics combined with newly developed techniques and protocols to study protein structure refinement and interactions of amino-acid analog pairs within lipid membranes are studied. A robust protocol for structure refinement of proteins from a given homologous model is designed and optimized that...
Show moreIn this dissertation, we have developed novel computational techniques that have been effectively utilized to extend our knowledge of proteins and lipid membrane systems. Application of molecular dynamics combined with newly developed techniques and protocols to study protein structure refinement and interactions of amino-acid analog pairs within lipid membranes are studied. A robust protocol for structure refinement of proteins from a given homologous model is designed and optimized that uses restrained molecular dynamics followed by optimal subset selection and structure averaging. This protocol is tested on CASP8 and CASP9 targets, and later successfully applied to CASP10 in blind prediction manner.In order to understand physical characteristics of peptide interactions embedded in bilayer membrane, we have used umbrella sampling technique with model amino acid side-chain analog pairs to study their association free energy while placed in membrane bilayer. As a result of convergence issues observed in such simulations due to bilayer deformation, a novel enhanced sampling technique is developed which biases the density of water in a cylinder, thereby effectively imposing bilayer deformation. Applying this method to a DPPC bilayer, we were able to study free energy of pore formation in membrane bilayers, and showed that while the undergone mechanism is different from currently existing methods, the mechanism by our proposed method is closer to the natural pore formation mechanism.
Show less