Abstract: Performance of any continuous speech recognition system is highly dependent on performance of the acoustic models. Generally, development of the robust spoken language technology relies on the availability of large amounts of data. Common way to cope with little data for training each state of Markov models is treebased state tying. This tying method applies contextual questions to tie states. Manual procedure for question generation suffers from human errors and is time consuming. Various automatically generated questions are used to construct decision tree. There are three approaches to generate questions to construct HMMs based on decision tree. One approach is based on misrecognized phonemes, another approach basically uses feature table and the other is based on state distributions corresponding to context-independent subword units. In this paper, all these methods of automatic question generation are applied to the decision tree on FARSDAT corpus in Persian language and their results are compared with those of manually generated questions. The results show that automatically generated questions yield much better results and can replace manually generated questions in Persian language.
Abstract: As a popular rank-reduced vector space approach,
Latent Semantic Indexing (LSI) has been used in information
retrieval and other applications. In this paper, an LSI-based content
vector model for text classification is presented, which constructs
multiple augmented category LSI spaces and classifies text by their
content. The model integrates the class discriminative information
from the training data and is equipped with several pertinent feature
selection and text classification algorithms. The proposed classifier
has been applied to email classification and its experiments on a
benchmark spam testing corpus (PU1) have shown that the approach
represents a competitive alternative to other email classifiers based
on the well-known SVM and naïve Bayes algorithms.
Abstract: In the paper, the energetic features of the loaded gait
are newly analyzed depending on the trunk flexion change. To
investigate the loaded gait, walking experiments are performed for five
subjects and, the ground reaction forces and kinematic data are
measured. Based on these information, we compute the impulse,
momentum and mechanical works done on the center of body mass,
through the trunk flexion change. As a result, it is shown that the trunk
flexion change does not affect the impulses and momentums during
the step-to-step transition as well. However, the direction of the
pre-collision momentum does change depending on the trunk flexion
change, which is degenerated just after (or during) the collision period.
Abstract: Many artificial intelligence (AI) techniques are inspired
by problem-solving strategies found in nature. Robustness is a key
feature in many natural systems. This paper studies robustness in
artificial neural networks (ANNs) and proposes several novel, nature
inspired ANN architectures. The paper includes encouraging results
from experimental studies on these networks showing increased
robustness.
Abstract: A manufacturing feature can be defined simply as a
geometric shape and its manufacturing information to create the shape.
In a feature-based process planning system, feature library plays an
important role in the extraction of manufacturing features with their
proper manufacturing information. However, to manage the
manufacturing information flexibly, it is important to build a feature
library that is easy to modify. In this paper, a Wiki-based feature
library is proposed.
Abstract: Recent years have witnessed the rapid development of
the Internet and telecommunication techniques. Information security
is becoming more and more important. Applications such as covert
communication, copyright protection, etc, stimulate the research of
information hiding techniques. Traditionally, encryption is used to
realize the communication security. However, important information
is not protected once decoded. Steganography is the art and science
of communicating in a way which hides the existence of the communication.
Important information is firstly hidden in a host data, such
as digital image, video or audio, etc, and then transmitted secretly
to the receiver.In this paper a data hiding model with high security
features combining both cryptography using finite state sequential
machine and image based steganography technique for communicating
information more securely between two locations is proposed.
The authors incorporated the idea of secret key for authentication
at both ends in order to achieve high level of security. Before the
embedding operation the secret information has been encrypted with
the help of finite-state sequential machine and segmented in different
parts. The cover image is also segmented in different objects through
normalized cut.Each part of the encoded secret information has been
embedded with the help of a novel image steganographic method
(PMM) on different cuts of the cover image to form different stego
objects. Finally stego image is formed by combining different stego
objects and transmit to the receiver side. At the receiving end different
opposite processes should run to get the back the original secret
message.
Abstract: Color image segmentation plays an important role in
computer vision and image processing areas. In this paper, the
features of Volterra filter are utilized for color image segmentation.
The discrete Volterra filter exhibits both linear and nonlinear
characteristics. The linear part smoothes the image features in
uniform gray zones and is used for getting a gross representation of
objects of interest. The nonlinear term compensates for the blurring
due to the linear term and preserves the edges which are mainly used
to distinguish the various objects. The truncated quadratic Volterra
filters are mainly used for edge preserving along with Gaussian noise
cancellation. In our approach, the segmentation is based on K-means
clustering algorithm in HSI space. Both the hue and the intensity
components are fully utilized. For hue clustering, the special cyclic
property of the hue component is taken into consideration. The
experimental results show that the proposed technique segments the
color image while preserving significant features and removing noise
effects.
Abstract: Semiconductor nanomaterials like TiO2 nanoparticles
(TiO2-NPs) approximately less than 100 nm in diameter have become
a new generation of advanced materials due to their novel and
interesting optical, dielectric, and photo-catalytic properties. With the
increasing use of NPs in commerce, to date few studies have
investigated the toxicological and environmental effects of NPs.
Motivated by the importance of TiO2-NPs that may contribute to the
cancer research field especially from the treatment prospective
together with the fractal analysis technique, we have investigated the
effect of TiO2-NPs on colony morphology in the dark condition
using fractal dimension as a key morphological characterization
parameter. The aim of this work is mainly to investigate the cytotoxic
effects of TiO2-NPs in the dark on the growth of human cervical
carcinoma (HeLa) cell colonies from morphological aspect. The in
vitro studies were carried out together with the image processing
technique and fractal analysis. It was found that, these colonies were
abnormal in shape and size. Moreover, the size of the control
colonies appeared to be larger than those of the treated group. The
mean Df +/- SEM of the colonies in untreated cultures was
1.085±0.019, N= 25, while that of the cultures treated with TiO2-NPs
was 1.287±0.045. It was found that the circularity of the control
group (0.401±0.071) is higher than that of the treated group
(0.103±0.042). The same tendency was found in the diameter
parameters which are 1161.30±219.56 μm and 852.28±206.50 μm
for the control and treated group respectively. Possible explanation of
the results was discussed, though more works need to be done in
terms of the for mechanism aspects. Finally, our results indicate that
fractal dimension can serve as a useful feature, by itself or in
conjunction with other shape features, in the classification of cancer
colonies.
Abstract: The various applications of VLSI circuits in highperformance
computing, telecommunications, and consumer
electronics has been expanding progressively, and at a very hasty
pace. This paper describes a new model for partitioning a circuit
using DBSCAN and fuzzy ARTMAP neural network. The first step
is concerned with feature extraction, where we had make use
DBSCAN algorithm. The second step is the classification and is
composed of a fuzzy ARTMAP neural network. The performance of
both approaches is compared using benchmark data provided by
MCNC standard cell placement benchmark netlists. Analysis of the
investigational results proved that the fuzzy ARTMAP with
DBSCAN model achieves greater performance then only fuzzy
ARTMAP in recognizing sub-circuits with lowest amount of
interconnections between them The recognition rate using fuzzy
ARTMAP with DBSCAN is 97.7% compared to only fuzzy
ARTMAP.
Abstract: Sensor Network are emerging as a new tool for
important application in diverse fields like military surveillance,
habitat monitoring, weather, home electrical appliances and others.
Technically, sensor network nodes are limited in respect to energy
supply, computational capacity and communication bandwidth. In
order to prolong the lifetime of the sensor nodes, designing efficient
routing protocol is very critical. In this paper, we illustrate the
existing routing protocol for wireless sensor network using data
centric approach and present performance analysis of these protocols.
The paper focuses in the performance analysis of specific protocol
namely Directed Diffusion and SPIN. This analysis reveals that the
energy usage is important features which need to be taken into
consideration while designing routing protocol for wireless sensor
network.
Abstract: This paper describes WiPoD (Wireless Position
Detector) which is a pure software based location determination and
tracking (positioning) system. It uses empirical signal strength measurements from different wireless access points for mobile user
positioning. It is designed to determine the location of users having
802.11 enabled mobile devices in an 802.11 WLAN infrastructure
and track them in real time. WiPoD is the first main module in our
LBS (Location Based Services) framework. We tested K-Nearest
Neighbor and Triangulation algorithms to estimate the position of a
mobile user. We also give the analysis results of these algorithms for
real time operations. In this paper, we propose a supportable, i.e.
understandable, maintainable, scalable and portable wireless
positioning system architecture for an LBS framework. The WiPoD
software has a multithreaded structure and was designed and implemented with paying attention to supportability features and real-time constraints and using object oriented design principles. We also describe the real-time software design issues of a wireless positioning system which will be part of an LBS framework.
Abstract: Image clustering is a process of grouping images
based on their similarity. The image clustering usually uses the color
component, texture, edge, shape, or mixture of two components, etc.
This research aims to explore image clustering using color
composition. In order to complete this image clustering, three main
components should be considered, which are color space, image
representation (feature extraction), and clustering method itself. We
aim to explore which composition of these factors will produce the
best clustering results by combining various techniques from the
three components. The color spaces use RGB, HSV, and L*a*b*
method. The image representations use Histogram and Gaussian
Mixture Model (GMM), whereas the clustering methods use KMeans
and Agglomerative Hierarchical Clustering algorithm. The
results of the experiment show that GMM representation is better
combined with RGB and L*a*b* color space, whereas Histogram is
better combined with HSV. The experiments also show that K-Means
is better than Agglomerative Hierarchical for images clustering.
Abstract: Real world Speaker Identification (SI) application
differs from ideal or laboratory conditions causing perturbations that
leads to a mismatch between the training and testing environment
and degrade the performance drastically. Many strategies have been
adopted to cope with acoustical degradation; wavelet based Bayesian
marginal model is one of them. But Bayesian marginal models
cannot model the inter-scale statistical dependencies of different
wavelet scales. Simple nonlinear estimators for wavelet based
denoising assume that the wavelet coefficients in different scales are
independent in nature. However wavelet coefficients have significant
inter-scale dependency. This paper enhances this inter-scale
dependency property by a Circularly Symmetric Probability Density
Function (CS-PDF) related to the family of Spherically Invariant
Random Processes (SIRPs) in Log Gabor Wavelet (LGW) domain
and corresponding joint shrinkage estimator is derived by Maximum
a Posteriori (MAP) estimator. A framework is proposed based on
these to denoise speech signal for automatic speaker identification
problems. The robustness of the proposed framework is tested for
Text Independent Speaker Identification application on 100 speakers
of POLYCOST and 100 speakers of YOHO speech database in three
different noise environments. Experimental results show that the
proposed estimator yields a higher improvement in identification
accuracy compared to other estimators on popular Gaussian Mixture
Model (GMM) based speaker model and Mel-Frequency Cepstral
Coefficient (MFCC) features.
Abstract: In this paper the features of multiple material gate
silicon-on-insulator MOSFETs are presented and compared with
single material gate silicon-on-insulator MOSFET structures. The
results indicate that the multiple material gate structures reduce short
channel effects such as drain induce barrier lowering, hot electron
effect and better current characteristics in comparison with single
material structures
Abstract: In this paper, a novel corner detection method is
presented to stably extract geometrically important corners.
Intensity-based corner detectors such as the Harris corner can detect
corners in noisy environments but has inaccurate corner position and
misses the corners of obtuse angles. Edge-based corner detectors such
as Curvature Scale Space can detect structural corners but show
unstable corner detection due to incomplete edge detection in noisy
environments. The proposed image-based direct curvature estimation
can overcome limitations in both inaccurate structural corner detection
of the Harris corner detector (intensity-based) and the unstable corner
detection of Curvature Scale Space caused by incomplete edge
detection. Various experimental results validate the robustness of the
proposed method.
Abstract: In this paper we propose a novel approach for ascertaining human identity based on fusion of profile face and gait biometric cues The identification approach based on feature learning in PCA-LDA subspace, and classification using multivariate Bayesian classifiers allows significant improvement in recognition accuracy for low resolution surveillance video scenarios. The experimental evaluation of the proposed identification scheme on a publicly available database [2] showed that the fusion of face and gait cues in joint PCA-LDA space turns out to be a powerful method for capturing the inherent multimodality in walking gait patterns, and at the same time discriminating the person identity..
Abstract: The paper shows some ability to manage two-phase
flows arising from the use of unsteady effects. In one case, we
consider the condition of fragmentation of the interface between the
two components leads to the intensification of mixing. The problem
is solved when the temporal and linear scale are small for the
appearance of the developed mixing layer. Showing that exist such
conditions for unsteady flow velocity at the surface of the channel,
which will lead to the creation and fragmentation of vortices at Re
numbers of order unity. Also showing that the Re is not a criterion of
similarity for this type of flows, but we can introduce a criterion that
depends on both the Re, and the frequency splitting of the vortices. It
turned out that feature of this situation is that streamlines behave
stable, and if we analyze the behavior of the interface between the
components it satisfies all the properties of unstable flows. The other
problem we consider the behavior of solid impurities in the extensive
system of channels. Simulated unsteady periodic flow modeled
breaths. Consider the behavior of the particles along the trajectories.
It is shown that, depending on the mass and diameter of the particles,
they can be collected in a caustic on the channel walls, stop in a
certain place or fly back. Of interest is the distribution of particle
velocity in frequency. It turned out that by choosing a behavior of the
velocity field of the carrier gas can affect the trajectory of individual
particles including force them to fly back.
Abstract: The recognition of human faces, especially those with
different orientations is a challenging and important problem in image
analysis and classification. This paper proposes an effective scheme
for rotation invariant face recognition using Log-Polar Transform and
Discrete Cosine Transform combined features. The rotation invariant
feature extraction for a given face image involves applying the logpolar
transform to eliminate the rotation effect and to produce a row
shifted log-polar image. The discrete cosine transform is then applied
to eliminate the row shift effect and to generate the low-dimensional
feature vector. A PSO-based feature selection algorithm is utilized to
search the feature vector space for the optimal feature subset.
Evolution is driven by a fitness function defined in terms of
maximizing the between-class separation (scatter index).
Experimental results, based on the ORL face database using testing
data sets for images with different orientations; show that the
proposed system outperforms other face recognition methods. The
overall recognition rate for the rotated test images being 97%,
demonstrating that the extracted feature vector is an effective rotation
invariant feature set with minimal set of selected features.
Abstract: In this paper a new approach to face recognition is presented that achieves double dimension reduction making the system computationally efficient with better recognition results. In pattern recognition techniques, discriminative information of image increases with increase in resolution to a certain extent, consequently face recognition results improve with increase in face image resolution and levels off when arriving at a certain resolution level. In the proposed model of face recognition, first image decimation algorithm is applied on face image for dimension reduction to a certain resolution level which provides best recognition results. Due to better computational speed and feature extraction potential of Discrete Cosine Transform (DCT) it is applied on face image. A subset of coefficients of DCT from low to mid frequencies that represent the face adequately and provides best recognition results is retained. A trade of between decimation factor, number of DCT coefficients retained and recognition rate with minimum computation is obtained. Preprocessing of the image is carried out to increase its robustness against variations in poses and illumination level. This new model has been tested on different databases which include ORL database, Yale database and a color database. The proposed technique has performed much better compared to other techniques. The significance of the model is two fold: (1) dimension reduction up to an effective and suitable face image resolution (2) appropriate DCT coefficients are retained to achieve best recognition results with varying image poses, intensity and illumination level.
Abstract: To achieve accurate and precise results of finite
element analysis (FEA) of bones, it is important to represent the
load/boundary conditions as identical as possible to the human body
such as the bone properties, the type and force of the muscles, the
contact force of the joints, and the location of the muscle attachment.
In this study, the difference in the Von-Mises stress and the total
deformation was compared by classifying them into Case 1, which
shows the actual anatomical form of the muscle attached to the femur
when the same muscle force was applied, and Case 2, which gives a
simplified representation of the attached location. An inverse
dynamical musculoskeletal model was simulated using data from an
actual walking experiment to complement the accuracy of the
muscular force, the input value of FEA. The FEA method using the
results of the muscular force that were calculated through the
simulation showed that the maximum Von-Mises stress and the
maximum total deformation in Case 2 were underestimated by 8.42%
and 6.29%, respectively, compared to Case 1. The torsion energy and
bending moment at each location of the femur occurred via the stress
ingredient. Due to the geometrical/morphological feature of the femur
of having a long bone shape when the stress distribution is wide, as
shown in Case 1, a greater Von-Mises stress and total deformation are
expected from the sum of the stress ingredients. More accurate results
can be achieved only when the muscular strength and the attachment
location in the FEA of the bones and the attachment form are the same
as those in the actual anatomical condition under the various moving
conditions of the human body.