Sharif Digital Repository

Multiple object tracking refers to the detection and following of target object classes in video sequences. In this task, all objects belonging to the target classes in the video are detected simultaneously in each frame, and a unique ID is assigned to each of them throughout the video. In recent years, the use of graph neural networks for solving this problem has received significant attention because these models are suitable tools for discovering and improving the relationships between objects in the scene, which can greatly assist in better object pairing. However, there are various challenges to using graph neural networks, the most important of which is the limitation of input graph... 

Performance-based design (PBD) has been accepted as one of the most reliable design methods in the past few decades. PBD can overcome inherent deficiencies associated with force-based design (FBD) and has advantages over this traditional method. Various PBD methods have been developed, which Displacement-based design (DBD) has been considered as one of the effective design approaches of the PBD method. In this study, the Displacement coefficient method (DCM) is developed for stub-type integral abutment bridges (IABs). For this purpose, two coefficients are proposed to determine the target displacement of these types of bridges. Validation of this method by nonlinear time-history analysis... 

Complex networks have recently become a very noticeable subject because of their remarkable theory on real networks such as rumor spreading. This thesis contributes to a prevention strategy which uses a maximal clique algorithm to find overlapping communities and also uses inverse targeting immunization strategy to immune these communities. By the using maximal clique algorithm combined with inverse targeting strategy we have achieved that the TIK method (targeted immunization using k-clique percolation) obtains 10% of better immunization than HD (highest degree) method which is one of the best methods in targeted immunization strategies. The TIK also demonstrates that rumor spreading is... 

Distance Metric Learning algorithms have been widely used in Machine Learning methods recently. In these algorithms a distance function between objecs (data points) is learned based on their labels or similarity and dissimilarity constraints. Recent works have shown that a good precision is obtained in classification or clustering methods which use these functions. Since in the current systems many of data points do not exist at the beginning and are added to the training set as the algorithm is run, online methods are needed to update learned metric due to new data.
In this thesis, we proposed a new online distance metric learning method that has higher performance than existing... 

Since the first observed case in December 2019, millions of people have been infected by Coronavirus disease 2019 (COVID-¬19), which is caused by severe acute respiratory syndrome coronavirus 2 (SARS¬-CoV-¬2). Located on the viral membrane, trimeric spike glycoprotein is one of the most crucial parts of the virus’ structure as it is responsible for the attachment and entry of the virus to host cells. The spike glycoprotein undergoes hinge-like conformational movements and its receptor¬binding domains (RBD) can be in either an accessible (up) state or an inaccessible (down) state to the host cell receptors, such as ACE2. Having performed Molecular Dynamics Simulations and Targeted Molecular... 

In this thesis, coupled trajectory and attitude control of an exoatmospheric interceptor is investigated. The interceptor is considered to track and intercept orbital targets in its terminal phase using an infrared strapdown seeker. Attitude control system uses its thrusters in order to keep permanently the detector toward the target in both elevation and azimuth directions. Detector is attached to the body. Divert thrusters generate acceleration perpendicular to the longitudinal axis of the interceptor to correct the trajectory. It is assumed that the interceptor has mass asymmetries. Transltional and rotational dynamics are coupled. Therefore, in order to improve the performance of... 

The frontend stalls caused by instruction cache and branch target buffer (BTB) misses are a well-known source of performance degradation in server processors.To address this limitation, a myriad of hardware prefetchers are proposed. While they can effectively eliminate lots of misses and increase performance, they are impractical solutions due to some shortcomings. In this study, we are looking for effective and practical solutions to address these limitations.Since the considerable fraction of instruction cache misses is sequential misses,sequential prefetchers like next-line prefetcher are simple and effective solutions to remove sequential misses that are used in modern server processors.... 

Drug-target binding affinity prediction is one of the most important and vital part of drug discovery. The computational methods to predict binfing affinity is a standing challenge in drug discovery. State-of-the-art models are usually based on supervised machine learning with known label information. It is expensive and time-consuming to collect labeled data. This thesis proposes a semi-supervised model based on convolutional GAN (Generative adversarial networks). The model consists of two Gans and Two CNN blocks for feature extraction and fully connected layers for prediction. Gan can learn protein and drug features from unlabeled data. We evaluate the performance of our method using four... 

A drug can be defined as a substance made to prevent disease, cure a specific symptom, relieve pain, and reduce anomalies in the body. The process of drug designing is so laborious, complex, costly, and time-consuming that chance of failure during the lab experiment stages is high. These challenges have persuaded researchers to find new usage for existing drugs, referred to as drug repurposing, with the main advantage of reducing cost, risk, and time. To this aim, computational methods have been applied to discover hidden pharmaceutical capabilities of drugs in terms of predicting whether a particular drug can interact with a particular protein.Graph Neural Networks (GNNs) have recently... 

Target localization in multiple-input multiple-output (MIMO) radar systems has been an active research topic during the recent years. Generally speaking, MIMO radars can be classified, based on the antennas geometry, into colocated and distributed architectures.The former provides waveform diversity and the latter makes use of spatial diversity provided by widely separated configuration of antennas, both of which improve localization and detection performance. In this study, we focus on the latter case. In this thesis, we investigate the localization problem in two parts, i.e., positioning based on bistatic time delay measurements and hybrid localization using combination of time delay,... 

Localization of radio, acoustic and vibration wave’s sources by passive systems has many applications in positioning systems, navigation systems, wireless sensor networks, defense, security, and geophysics. RSS, AOA, TDOA, and FDOA are some of the techniques available for passive localization. The performance of TDOA/FDOA techniques does not degrade with distance and multipath, but it may suffer from poor performance for narrowband signals. On the other hand, FDOA technique requires narrowband signals. Thus, the combination of TDOA and FDOA can be suitable for a much wider range of sources. In addition, the TDOA/FDOA method can provide more accurate interference source localization compared... 

Research on target localization in multiple-input multiple-output (MIMO) radars, has seen significant growth in recent years. In general, localization methods in these radars can be classified as direct and indirect methods. In the former, the cinematic specifications of target is directly extracted from the received signals, whereas in the indirect methods, different measurements are calculated from the received signals and then, the target position is estimated using the acquired measurements. In this thesis, by concentrating on indirect target localization methods and by taking practical problems, such as uncertainties in sensors position and velocity, asynchronous sensor clocks and... 

Range Profile (RP) is known as one of the most important tools for radar target recognition. The main problem with range profile for radar target recognition is its sensitivity to aspect angle. To overcome this problem, range profiles are assumed to have the same statistical characteristics in small frames of aspect angles or context dependent models such as HMM may be used. All methods presented, seems to have some shortcomings to offer a model based on physical circumstances of target maneuver. For example in such models, consecutive samples of RPs in an aspect frame are assumed to be statistically independent with the same distribution (IID). Here we propose dynamic system (DS) for... 

In recent years, application of nanotechnology in medicine is growing rapidly, specially, in the area of drug delivery. One of most import applications of nanotechnology is the using of nanoparticles as a carrier in targeted drug delivery systems. Simulation of drug delivery and prediction of drug release are developing in experimental and industrial areas. Mathematical modeling is used in drug delivery systems because they are time and cost saving and also can be used to predict drug behavior. The aim of this research, is to provide a framework for designing nano drug carriers. To this end, uptake of nanoparticles into different types of cells with different characteristics, in different... 

A brief overview of the proposed research, its aims, engineering relevance and applications; as well as expected achievements compared to the present state of the art:The plastic hinge length and ultimate drift calculations of frames are required in the performance based design of reinforce concrete structure. FRP-confinement increases the column’s shear resistance and ductility. FRP-confinement prevents concrete cover from spalling and increases the inelastic deformability of concrete in the potential plastic hinge region. GU et al. 2012 [24] studied the existing experimental results of confinement effects on the plastic hinge length. Mortezaei and Ronagh, 2012 [25], studied the plastic... 

Today, current methods are used for treatment of cancers. Patient experiences a period of hard treatment during destruction of both healthy and diseased cells. Researchers are following stimulus- response drug delivery systems that they control time and position of drug release. Meanwhile, micellar nanoparticles are important, because it can transport the hydrophobic anticancer drugs from body's physiological barriers. On the other hand, tumour tissue has low pH and high temperature as compared to health tissue. Therefore, using pH and thermosensitive polymer coating could be effective in triggered drug release. In the first study, micellar systems were synthesized which were responsive to... 

Mesoporous silica nanoparticles have broad application in drud delivery systems due to their porous structure, functionalization, biocompatibility, high surface area and pore volume. Neverthless, pure mesoporous silica nanoparticles without functionality were not smart material and could not release drug in triggered and controlled manner. For this reason, using smart polymeric coating would be considered. Polymer shells also provide colloidal stability, improved blood circulation lifetime and reduced toxicity which are crucial for efficient in vivo drug delivery. Inflammatory and tumor tissue have low pH and high temperature as compared to health tissue. Therefore, using pH and... 

In this thesis a hybrid guidance law has been designed for a ground-based system to intercept an exoatmospheric target. In this regard a three stage interceptor with a kinetic kill vehicle has been proposed, the design parameters of which are determined according to a conceptual design procedure. The guidance law consists three phases namely boost, midcourse and terminal phases. In boost phase, a ballistic guidance law steers the first and the second stages of the interceptor ahead of the predeicted intercept point. In the midcourse phase, the third stage of the interceptor corrects the engagement plane while it also provides the head on interception condition for the kinetic kill vehicle.... 

This study proposes an efficient risk-targeted framework for integrated optimal seismic design and quantifying target safety of steel Special Moment-Resisting Frames (steel SMRFs) with and without various passive damping devices such as Buckling-Restrained Braces (BRBs), Friction Dampers (FDs), and Fluid Viscous Dampers (FVDs). This framework provides an automated design procedure formulated as a constrained single-objective optimization problem to achieve the minimum Life Cycle Cost (LCC) within the Particle Swarm Optimization (PSO) algorithm and can be the basis of current seismic codes for target safety calibration. LCC includes initial construction cost and lifetime seismic losses such... 

In MTI radars, digital filters utilizing non-uniform sampling are used to remove the echo of stationary targets or the ones with minor Doppler shift. These targets usually are associated with significant reflection causing small, high-speed targets not to be detected. In this type of filters, the use of non-uniform sampling prevents the blind speed issue and causes the first blind speed to occur at higher frequencies. The performance of such filters is characterized by two criteria: Coefficients and periods. Therefore, coefficient optimization in filter design has a dominant role in maximizing clutter rejection. Also, period optimization is used to prevent generating deep nulls in the...