Sharif Digital Repository

Data stream prediction is challenging when concepts drift, processing time, and memory constraints come into account. Concept drift refers to changes in data distribution over time that reduces prediction systems’ accuracy. We present a method for handling concept drift with a domain adaptation approach (CDDA) in a data stream. The proposed method passively deals with the concept drift by using the domain adaptation approaches with multiple sources while reducing the model execution time and memory consumption. We introduce two variants of CDDA to transfer the information in the multi-source windows to the target window: weighted multi-source CDDA and multi-source feature alignment CDDA.... 

In large databases, availability of labeled training data is mostly prohibitive in classification. Semi-supervised algorithms are employed to tackle the lack of labeled training data problem. Video databases are the epitome for such a scenario; that is why semi-supervised learning has found its niche in it. Graph-based methods are a promising platform for semi-supervised video classification. Based on the multiview characteristic of video data, different features have been proposed (such as SIFT, STIP and MFCC) which can be utilized to build a graph. In this paper, we have proposed a new classification method which fuses the results of manifold regularization over different graphs. Our... 

A robotic system was designed and developed to perform the camera handling task during laparoscopic surgery. The system employs an effective low cost mechanism, with a minimum number of actuated DOFs, enabling spherical movement around a remote centre of motion positioned at the the insertion point of the laparoscopic stem. Kinematic analysis showed a high manipulability measure for the system, with the left/right movements directly governed by rotation of the first rotary actuator, and zoom and up/down movements by the simultaneous motions of the linear and second rotary actuators. A prototype of the robot was developed for practical use in an operating room environment. Hands-free operator... 

Porous gelatin-chitosan microcarriers (MCs) with the size of 350 ± 50 μm were fabricated with blends of different gelatin/chitosan (G/C) weight ratio using an electrospraying technique. Response surface methodology (RSM) was used to study the quantitative influence of process parameters, including blend ratio, voltage, and syringe pump flow rate, on MCs diameter and density. In the following, MCs of the same diameter and different G/C weight ratio (1, 2, and 3) were fabricated and their porosity and biocompatibility were investigated via SEM images and MTT assay, respectively. The results showed that mesenchymal stem cells (MSCs) could attach, proliferate, and spread on fabricated porous MCs... 

To get a more accurate model for simulation of single electron transistors (SETs), we have proposed a new macromodel that includes the ability of electron tunneling time calculation. In our proposed model, we have modified the previous models and applied some basic corrections to their formulas. In addition, we have added a switched capacitor circuit, as a quantizer, to calculate the electron tunneling time. We used HSPICE for high-speed simulation and observed that the simulation results obtained from our model matched more closely with that of SIMON 2.0. We also could evaluate the time of electron tunneling through the barrier by using the quantizer. Clearly, our macro-model gives more... 

Voluminous amount of videos in today’s world has made the subject of objective (or semi-objective) classification of videos to be very popular. Among the various descriptors used for video classification, SIFT and LIFT can lead to highly accurate classifiers. But, SIFT descriptor does not consider video motion and LIFT is time-consuming. In this paper, a robust descriptor for semi-supervised classification based on video content is proposed. It holds the benefits of LIFT and SIFT descriptors and overcomes their shortcomings to some extent. For extracting this descriptor, the SIFT descriptor is first used and the motion of the extracted keypoints are then employed to improve the accuracy of... 

To get a more accurate model for the simulation of single electron transistors (SETs), we propose a new macro-model that includes an electron tunneling time calculation. In our proposed model, we have modified the previous models and have applied some basic corrections to the formulas. In addition, we have added a switched capacitor circuit, as a quantizer, to calculate the electron tunneling time. We used HSPICE for a high-speed simulation and observed that the simulation results obtained from our model match more closely with that of SIMON 2.0. We also could evaluate the time of electron tunneling through the barrier by using the quantizer. Clearly, our macro-model gives more accurate... 

In many sensor fusion algorithms, the vision based RANdom Sample Consensus (RANSAC) method is used for estimating motion parameters for autonomous robots. Usually such algorithms estimate both translation and rotation parameters together which makes them inefficient solutions for merely rotation estimation purposes. This paper presents a novel 3-point RANSAC algorithm for estimating only the rotation parameters between two camera frames which can be utilized as a high rate source of information for a camera-IMU sensor fusion system. The main advantage of our proposed approach is that it performs less computations and requires fewer iterations for achieving the best result. Despite many... 

Although the accuracy of MEMS gyroscopes has been extremely improved, in some aspects, such as stability of bias, they still suffer from some big error sources, like run-to-run bias, which determines the sensor price but is not negligible even inexpensive sensors. Due to the fact that run-to-run bias is a kind of stochastic parameter, it has to be measured by utilizing online methods. Utilizing a novel, fast and efficient vision-based rotation estimation algorithm for ground vehicles, we have developed a visual gyroscope that is used in our sensor fusion system, in order to estimate run-to-run bias of the MEMS gyroscope, accurately. Comparing with similar approaches that use GPS, odometer,... 

The 2D hypersonic real gas flow has been analyzed on an adaptive unstructured grid using Roe's Flux Difference Splitting and AUSM schemes. In high temperature and hypersonic regime, the flow is extremely compressible and ideal gas assumption is not valid. In fact in these flows, due to changes in the flow properties, composition of fluid elements will also change. To solve steady and unsteady 2D Euler' equations for real gases, assumption of a general equation of state for real gases in equilibrium is considered. We use an unstructured Delaunay triangulation and adapt it in high gradient areas. Results are compared with known numerical and exact solutions. The scheme is convergent, and... 

Numerical simulation of resin transfer molding (RTM) is known as a useful method to analyze the process before the mold is actually built. In thick parts, the resin flow is no longer two-dimensional and must be simulated in a fully three-dimensional space. This article presents numerical simulations of three-dimensional non-isothermal mold filling of the RTM process. The control volume/finite element method (CV/FEM) is used in this study. Numerical formulation for resin flow is based on the concept of nodal partial saturation at the flow front. This approach permits to include a transient term in the working equation, removing the need for calculation of time step to track the flow front in... 

A process cycle of resin transfer molding (RTM) consists of two sequential stages, i.e. filling and curing stages. These two stages are interrelated in non-isothermal processes so that the curing stage is dominated by the resin flow as well as temperature and conversion distributions during the filling stage. Therefore, it is necessary to take into account both filling and curing stages to analyze the process cycle accurately. In this paper, a full three-dimensional process cycle simulation of RTM is performed. Full three-dimensional analysis is necessary for thick parts or parts having complex shape. A computer code is developed based on the control volume/finite element method (CV/FEM).... 

In this study the authors present an in-depth study, analysis and discussion on maximum likelihood (ML)-based receiver for a typical spectrally encoded spread-time CDMA in the presence of multiple narrowband interference (NBI) signals in an additive white Gaussian noise channel. Furthermore, the authors demonstrate that by combining useful properties of ML-based receiver and three-level codes, that is, codes with values of -1, 0, +1, the authors can introduce a new strategy in which superior performance with respect to previous receiver structure based on two-level codes, that is, codes with values of -1, +1, can be attained. With the help of an example, the authors drive, first, the...