Sharif Digital Repository

The advent of digital age with the Internet revolution has made it extremely convenient for users to access, create, process, copy, or exchange multimedia data. This has created an urgent need for protecting intellectual property in both the digital and the print media. Digital watermarking is a suitable way to do this. In this technology, some hidden information called watermark are embedded into host signal and extracted to confirm copyright protection. However, the watermark should be embedded in the host in such a way that the attacks could not destroy it. Print and scan is a popular attack that is applied on digital images. This attack has complex nature and can be implemented easily.... 

Although many mathematicians have searched on fractional calculus since many years ago, but its application in engineering, especially in modeling and control, does not have many antecedents. Because there are much freedom in choosing the order of differentiator and integrator in fractional calculus, it is possible to model the physical systems accurately. Also, there is the same reasoning in the control area. We can design such fractional controllers that have high performance. Among the fractional order controllers, the fractional order PID (FOPID) has been dealt with more than others and there are many methods to design it, such as Ziegler-Nichols, state space and method based on genetic... 

The LQG theory made its appearance in the fifties and sixties. It has now become one of the standard methods for compensator design. Despite the existing powerful tools in modern control, there seems to be few or no systematic approaches proposed for determining the weighting and root matrices which are the free parameters of the compensator. These matrices are often found by trial-and-error. Initially, this thesis presents an iterative algorithm for determining the compensator parameters. The algorithm is based upon making corrections to the singular values graphs in order to enhance closed-loop performance and robustness. This is how the traditional intuitive trial-and-error approach is... 

Up to now, compared with the comprehensive research for developing robust watermarking algorithms, no equal attention has been devoted to the proposition of benchmarks tailored to assess the watermark robustness. In addition, almost all the state of the art benchmarks only integrate a number of common image processing operations like geometrical transformations to remove watermarks. However, the quality of the processed image is often too degraded to permit further commercial exploitation. Moreover, to the best of our knowledge, the design of these tools does not take into account the statistical properties of the images and watermarks in the design of attacks. In spite of the significant... 

In complex human social activities, practical problems involve more prominent uncertainty, and deterministic approaches of classical methods become powerless. In this research, a new method for human group decision making is presented by using heterogeneous incomplete uncertain preference relations. The uncertain multiplicative preference relations, uncertain fuzzy preference relations, uncertain linguistic preference relations, intuitionistic fuzzy preference relations and Interval
preference sequence can be included in this method.Our new method consists of nine steps. In the first step, decision-makers preferences in the form of heterogeneous comparison matrixes are driven.In the... 

This thesis, argues on recently published methods for tuning of fractional PID controllers. Proposed algorithms provided tuning rules to achieve design objective such as, phase margin, gain cross-over frequency and zero derivative of the phase versus frequency at cross-over frequency. Third constraint imposed to reach high robustness to gain loop variation for controller. Tuning rules attained by consideration on the insensitivity of both stability and desired performance to the system parameter variations, but not on own controller. At the present paper, based on geometric approach, fragility analysis of the fractional-order of some types of PID setting for considering the robustness of the... 

Complex Networks are everywhere, many complex systems can be represented as networks, such as the power grid, the road network, the airline network and the Protein-protein interaction network, delivery and distribution networks, and telephone networks. A fundamental issue concerning complex networks is the robustness of the overall system to the failure of its constituent parts. The robustness of networks against failure, targeted attacks to individuals’ components, and the impact on the performance of the system has become an important issue for practical reasons in the last few years. The failures attacks or errors on networks are not limited to the deletion of vertex, for common... 

In recent years, neural networks have been widely used in most machine learning domains. However, it has been shown that these networks are vulnerable to adversarial examples. adversarial examples are small and imperceptible perturbations applied to the input which lead to producing wrong output and thus, fooling the network. This will become an important issue in security related applications of deep neural networks, such as self-driving cars and medical diagnostics. Since, in the wort-case scenario, even human lives could be threatened. Although, many works have focused on crafting adversarial examples for image data, only a few studies have been done on textual data due to the existing... 

Despite widespread applications and high performance of deep neural networks in the fields of computer vision, they have been shown to be vulnerable to adversarial examples. An adversarial example is a perturbated image that the magnitude of its difference with its corresponding natural image is small and yet given such example, the network produces incorrect output. In recent years, many approaches have been proposed to increase the robustness of DNNs against adversarial examples with adversarial training being proposed as the most effective defense measure. Approaches based on adversarial training try to increase the robustness of the network by training on the adversarial examples. One of... 

Question Answering (QA) systems have reached human-level accuracy; however, these systems are vulnerable to adversarial examples. Recently, adversarial attacks have been widely investigated in text classification. However, there have been few research efforts on this topic in QA systems. In this thesis our approach is improving the robustness of QA systems using deep neural networks. In this thesis, as the first proposed approach, the knowledge distillation method is introduced to create a student model to improve the robustness of QA systems. In this regard, the pre-trained BERT model was used as a teacher, and its impact on the robustness of the student models on the Adversarial SQuAD... 

Few-shot learning methods have achieved notable performance in recent years. However, fewshot learning in large-scale settings with hundreds of classes is still challenging. In this dissertation, we tackle the problems of large-scale few-shot learning by taking advantage of pre-trained foundation models. We recast the original problem in two levels with different granularity. At the coarse-grained level, we introduce a novel object recognition approach with robustness to sub-population shifts. At the fine-grained level, generative experts are designed for few-shot learning, specialized for different superclasses. A Bayesian schema is considered to combine coarse-grained information with... 

This dissertation proposes a probabilistic framework for optimal calibration of design provisions based on the minimization of lifecycle cost (LCC) and its uncertainty. Subsequently, this framework is utilized to determine the optimal robust design base shear coefficient of building structures, and propose a methodology for codifying it while preserving the current structure of the base shear equation. In the first part of the proposed methodology, various structures are designed with different seismic base shears. Then, at each site, their LCC comprising the construction costs and seismic losses is calculated. Various types of seismic loss considered in this study include the direct...