Loading...
Search for:
soltani-farani--ali
0.152 seconds
Total 2609 records
Detecting Speakers in a Telephone Conversation
, M.Sc. Thesis Sharif University of Technology ; Sameti, Hossein (Supervisor)
Abstract
The human speech signal conveys many levels of information ranging from phonetic content to speaker identity and even emotional status. This thesis deals with the task of open-set speaker identification (SI) from an unconstrained telephone conversation between two speakers. The goal is to find at most two speakers among a known set of target speakers that best match the voice samples of the input speech; the input voice samples are not constrained to the target speaker set. The uni-speaker problem is investigated first. The classic GMM-UBM system for text-independent SI and its adapted form are explored. The use of score-space information is advocated as a complementary source to the...
When pixels team up: Spatially weighted sparse coding for hyperspectral image classification
, Article IEEE Geoscience and Remote Sensing Letters ; Volume 12, Issue 1 , Jan , 2015 , Pages 107-111 ; 1545598X (ISSN) ; Rabiee, H. R ; Sharif University of Technology
Institute of Electrical and Electronics Engineers Inc
2015
Abstract
In this letter, a spatially weighted sparse unmixing approach is proposed as a front-end for hyperspectral image classification using a linear SVM. The idea is to partition the pixels of a hyperspectral image into a number of disjoint spatial neighborhoods. Since neighboring pixels are often composed of similar materials, their sparse codes are encouraged to have similar sparsity patterns. This is accomplished by means of a reweighted ℓ1 framework where it is assumed that fractional abundances of neighboring pixels are distributed according to a common Laplacian Scale Mixture (LSM) prior with a shared scale parameter. This shared parameter determines which endmembers contribute to the group...
Structured Sparse Representation for Machine Learning and Signal
Processing
,
Ph.D. Dissertation
Sharif University of Technology
;
Rabiei, Hamid Reza
(Supervisor)
Abstract
In this proposal we pursue structured sparse representations. Recent years have witnessed a tremendous growth in sparse modeling of natural signals. In this model a signal is represented as a linear combination of a few atoms from an often over-complete dictionary. The recent success of compressed sensing is intact due to the property that natural signals often admit sparse representations. Still, sparse representation in its simplest form sometimes fails to capture the intrinsic structure in natural signals. This structure may be embedded in the signal itself or in the relation between different signals of interest. The goal of this research proposal is to exploit the intrinsic structure in...
Spatial-aware dictionary learning for hyperspectral image classification
, Article IEEE Transactions on Geoscience and Remote Sensing ; Volume 53, Issue 1 , July , 2015 , Pages 527-541 ; 01962892 (ISSN) ; Rabiee, H. R ; Hosseini, S. A ; Sharif University of Technology
Institute of Electrical and Electronics Engineers Inc
2015
Abstract
This paper presents a structured dictionary-based model for hyperspectral data that incorporates both spectral and contextual characteristics of spectral samples. The idea is to partition the pixels of a hyperspectral image into a number of spatial neighborhoods called contextual groups and to model the pixels inside a group as members of a common subspace. That is, each pixel is represented using a linear combination of a few dictionary elements learned from the data, but since pixels inside a contextual group are often made up of the same materials, their linear combinations are constrained to use common elements from the dictionary. To this end, dictionary learning is carried out with a...
Patchwise joint sparse tracking with occlusion detection
, Article IEEE Transactions on Image Processing ; Vol. 23, Issue. 10 , 2014 , Pages. 4496-4510 ; ISSN: 10577149 ; Rabiee, H. R ; Soltani-Farani, A ; Khajenezhad, A ; Sharif University of Technology
2014
Abstract
This paper presents a robust tracking approach to handle challenges such as occlusion and appearance change. Here, the target is partitioned into a number of patches. Then, the appearance of each patch is modeled using a dictionary composed of corresponding target patches in previous frames. In each frame, the target is found among a set of candidates generated by a particle filter, via a likelihood measure that is shown to be proportional to the sum of patch-reconstruction errors of each candidate. Since the target's appearance often changes slowly in a video sequence, it is assumed that the target in the current frame and the best candidates of a small number of previous frames, belong to...
Classifying a stream of infinite concepts: A Bayesian non-parametric approach
, Article Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) ; Vol. 8724 LNAI, issue. PART 1 , 2014 , p. 1-16 ; Rabiee, H.R ; Hafez, H ; Soltani-Farani, A ; Sharif University of Technology
2014
Abstract
Classifying streams of data, for instance financial transactions or emails, is an essential element in applications such as online advertising and spam or fraud detection. The data stream is often large or even unbounded; furthermore, the stream is in many instances non-stationary. Therefore, an adaptive approach is required that can manage concept drift in an online fashion. This paper presents a probabilistic non-parametric generative model for stream classification that can handle concept drift efficiently and adjust its complexity over time. Unlike recent methods, the proposed model handles concept drift by adapting data-concept association without unnecessary i.i.d. assumption among the...
Filter-bank design based on dependencies between frequency components and phoneme characteristics
, Article European Signal Processing Conference, 29 August 2011 through 2 September 2011 ; Septembe , 2011 , Pages 2142-2145 ; 22195491 (ISSN) ; Sameti, H ; Tavanaei, A ; Soltani Farani, A ; Sharif University of Technology
2011
Abstract
Mel-frequency Cepstral coefficients are widely used for feature extraction in speech recognition systems. These features use Mel-scaled filters. A new filter-bank based on dependencies between frequency components and phoneme characteristics is proposed. F-ratio and mutual information are used for this purpose. A new filter-bank is designed in which frequency resolution of sub-band filters is inversely proportional to the computed dependency values. These new filterbank is used instead of Mel-scaled filters for feature extraction. A phoneme recognition experiment on FARSDAT Persian language database showed that features extracted using the proposed filter-bank reach higher accuracy (63.92%)...
Semi-spatiotemporal fMRI brain decoding
, Article Proceedings - 2013 3rd International Workshop on Pattern Recognition in Neuroimaging, PRNI 2013 ; 2013 , Pages 182-185 ; 9780769550619 (ISBN) ; Sheikhzadeh, H ; Rabiee, H. R ; Soltani Farani, A ; Sharif University of Technology
2013
Abstract
Functional behavior of the brain can be captured using functional Magnetic Resonance Imaging (fMRI). Even though fMRI signals have temporal and spatial structures, most studies have neglected the temporal structure when inferring mental states (brain decoding). This has two main side effects: 1. Degradation in brain decoding performance due to lack of temporal information in the model, 2. Inability to provide temporal interpretability. Few studies have targeted this issue but have had less success due to the burdening challenges related to high feature-to-instance ratio. In this study, a novel model for incorporating temporal information while maintaining a low feature-to-instance ratio, is...
A probabilistic joint sparse regression model for semisupervised hyperspectral unmixing
, Article IEEE Geoscience and Remote Sensing Letters ; Volume 14, Issue 5 , 2017 , Pages 592-596 ; 1545598X (ISSN) ; Rabiee, H. R ; Soltani Farani, A ; Zarezade, A ; Sharif University of Technology
Institute of Electrical and Electronics Engineers Inc
2017
Abstract
Semisupervised hyperspectral unmixing finds the ratio of spectral library members in the mixture of hyperspectral pixels to find the proportion of pure materials in a natural scene. The two main challenges are noise in observed spectral vectors and high mutual coherence of spectral libraries. To tackle these challenges, we propose a probabilistic sparse regression method for linear hyperspectral unmixing, which utilizes the implicit relations of neighboring pixels. We partition the hyperspectral image into rectangular patches. The sparse coefficients of pixels in each patch are assumed to be generated from a Laplacian scale mixture model with the same latent variables. These latent variables...
Peer-to-Peer Compressive Sensing for Network Monitoring
, Article IEEE Communications Letters ; Volume 19, Issue 1 , September , 2015 , Pages 38-41 ; 10897798 (ISSN) ; Rabiee, H. R ; Siyari, P ; Soltani Farani, A ; Khodadadi, A ; Sharif University of Technology
Institute of Electrical and Electronics Engineers Inc
2015
Abstract
Monitoring large-scale networks is a critical yet challenging task. Enormous number of nodes and links, limited power, and lack of direct access to the entire network are the most important difficulties. In applications such as network routing, where all nodes need to monitor the status of the entire network, the situation is even worse. In this letter, a collaborative model in which nodes pick up information from measurements generated by other nodes is proposed. Using this model, for the first time, an upper bound is derived for the number of measurements that each node must generate, such that the expected number of measurements observed by each node is sufficient to provide a global view...
Structural Health Monitoring of composite Plates with Piezoelectric WaferActive Sensors
, M.Sc. Thesis Sharif University of Technology ; Abedian, Ali (Supervisor)
Abstract
Cost-effective and reliable damage detection is critical for utilization of composite materials. This research investigates a wavelet-based approach for SHM and detection of through-thickness crack in different lengths and angles in [0/45/-45/90]s CF/EP composite plate utilizing ANSYS finite element method (FEM) software and digital signal processing technique utilizing MATLAB software. To this end, a dual-PZT actuation scheme is applied, aimed at generating the S0 mode at a square composite plate. This study contributes to the damage detection by developing an improved wavelet-based signal processing technique that enhances the visibility and interpretation of the Lamb wave signals related...
Partial Vocabulary Learning through Multimedia Input: An Analysis of the Effects of Captions and Subtitles
, M.Sc. Thesis Sharif University of Technology ; Jahangard, Ali (Supervisor)
Abstract
The current study aimed to investigate the differential effects of multimedia input with no subtitle, subtitle, or caption (L2 subtitle) on vocabulary retention while controlling for the possible effects of students' vocabulary size and language aptitude as covariates. A total of 81 undergraduate university students were assigned to the study from three intact groups at Sharif University of Technology. All of the participants were Persian speakers and were in the age range of 18 and 22 years old; homogeneity of the students regarding their language proficiency was also established. During two different sessions, students' vocabulary size and language aptitude were measured. The groups were...
Investigating Market Reaction to Asset Revaluation
, M.Sc. Thesis Sharif University of Technology ; Ebrahimnejad, Ali (Supervisor)
Abstract
In recent years, some listed companies have re-evaluated their assets, an event that has generally received market attention; In fact, the market seems to be reacting very positively to an accounting phenomenon. In this study, we seek to investigate the market reaction to the revaluation of listed companies in Iran. To do this, we first estimate the market reaction to the news of the revaluation of assets. We then examine the possible reasons for the market reaction to the revaluation. Asset revaluation can improve the firm's access to finance by improving some financial ratios; On the other hand, the increase in capital from the revaluation of firms is accompanied by the issuance of bonus...
Integrated Design of Cellular Manufacturing Systems
, M.Sc. Thesis Sharif University of Technology ; Houshmand, Mahmoud (Supervisor)
Abstract
Increasing competition in global market and different requirements of customers persuade manufacturers to improve the efficiency of their manufacturing activities. For this end cellular manufacturing systems can be introduced as an innovative and compromising practice for companies. Among the necessary decisions for a successful cellular manufacturing design and implementation, production routing selection (PRS), cell formation (CF), group layout (GL) and group scheduling (GS) are the four most important ones. This research investigates these decisions and proposes a new approach to integrate them. We use important production factors such as production volume, operation sequence, operational...
Investigation of Deformation Behavior and Stress Relaxation of Helical Springs
, M.Sc. Thesis Sharif University of Technology ; Karimi Taheri, Ali (Supervisor)
Abstract
Springs parts are widely used in many industries including the automative industries where the disability of the springs can cause damage. Spring forming processes induce residual stresses on the surface of spring. Residual stress is affected by various processes, such as heat treatment or shot peening. When a coiled spring is heat treated, depending on the temperature and time of heat treatment residual stresses are partially relieved, but they generally do not disappear.An important characteristic of spring steel is its resistance against stress relaxation at different temperatures and times. Stress relaxation in valve springs working at high temperatures causes some problems for...
A Nonlinear Layerwise Shell Finite Element for Delamination Analysis of Laminated Composite Structures under Large Deformation
, Ph.D. Dissertation Sharif University of Technology ; Hosseini Kordkheili, Ali (Supervisor)
Abstract
This thesis aims to develop a numerically efficient nonlinear layer-wise shell element formulation for delamination analysis of laminated composite shell structures. The element, in a mesoscale scheme, is formulated based on a zig-zag theory and features three translational degrees of freedom for each node on the mid-surface of the shell in addition with two rotational degrees of freedom for each layer. In this way, the displacement field is formulated via adapting the Mindlin-Reissner theory in each layer and an ordered second-order algorithm for finite rotations. To verify the proposed formulation, many popular benchmark problems for geometric nonlinear analysis of shell problems are...
Factors Affecting Index Funds Performance: The Case of Karafarin Index Fund
, M.Sc. Thesis Sharif University of Technology ; Ebrahimnejad, Ali (Supervisor) ; Talebian, Masoud (Supervisor)
Abstract
Index funds are one of the most vital financial institutions providing a wide range of values for different investors. Despite the importance of these funds, not only are index funds recently developed, but they also do not replicate their indices effectively. In this paper, we analyze the portfolio performance of the Karafarin Index Fund as an open-end fund for case study in various frameworks. First of all, the asset allocation and security selection attributions are assessed for capturing the capabilities to choose the suitable security at the right time. In addition, this assessment is done for the industry allocation and security selection attribution in the stock portfolio of the...
An Experimental Study of the Effects of Sweep Wing on the Boundary Layer of 2D Wing
, M.Sc. Thesis Sharif University of Technology ; Soltani, Mohammad Reza (Supervisor) ; Davari, Ali Reza (Co-Advisor)
Abstract
The behavior of boundary layer under the effect of the sweep angle is considered. The measurements were performed by a pitot tube rake. Three models with various sweep angels at angles of attack -2, 0 and 2 degree were tested. Both tip and root of all wings were closed by flat plates in order to prevent the flow to roll-up. However the flow field on the wing was still three-dimensional because of the sweep angle. The velocity on the wing has two components; longitudinal and lateral. The cross flow emerges due to the non-equilibrium of pressure and centrifugal forces. The velocity profiles showed that the magnitude of cross flow was stronger near the leading edge of the wing. The cross flow...
Fabrication and Mechanical Properties of Alumina- Zirconia- CNT Nanocomposite Parts
, M.Sc. Thesis Sharif University of Technology ; Faghihi Sani, Mohammad Ali (Supervisor) ; Shakhesi, Saeeid (Supervisor)
Abstract
Alumina as an engineering ceramic has special characteristics such as hardness, high abrasive, fatigue and corrosion resistance but suffers from low toughness. One effective method in improving toughness is adding second phases such as tetragonal zirconia and carbon nanotubes.
In this research, alumina- zirconia nanopowders has been prepared by solution combustion method and carbon nanotubes was added to it by two different ways: 1) in situ synthesis of CNTs by CVD 2) dispersion of CNTs by heterocoagulation method. In the case of in situ synthesis of CNTs in AZ matrix, the synthesis temperature and flow ratio of N2/C2H2 gases was investigated and optimum conditions were 800̊C and 300/30,...
In this research, alumina- zirconia nanopowders has been prepared by solution combustion method and carbon nanotubes was added to it by two different ways: 1) in situ synthesis of CNTs by CVD 2) dispersion of CNTs by heterocoagulation method. In the case of in situ synthesis of CNTs in AZ matrix, the synthesis temperature and flow ratio of N2/C2H2 gases was investigated and optimum conditions were 800̊C and 300/30,...
Utilization of Airborne Wind Energy in Iran
, M.Sc. Thesis Sharif University of Technology ; Vesaghi, Mohammad Ali (Supervisor) ; Soltani, Mohammad Reza (Supervisor)
Abstract
A Flying Electric Generator (FEG) can be a wind turbine which operates at high altitudes, where there is a continuous fast flow of air and converts wind energy to electrical power.The related field is an interdisciplinary area and is mostly related to the Aerospace Engineering and Energy Conversion. We focused on Iran, as one of the few lands in the northern hemisphere that has a high overall amount of high altitude wind energy for at least half of the year over it, according to the information of atlas of airborne wind energy in this work.So to install the FEGs in Iran, we had to find, places with a safe aerial space overhead which grant us reliable spots over country. These places found by...