Abstract: In this study, a high accuracy protein-protein interaction
prediction method is developed. The importance of the proposed
method is that it only uses sequence information of proteins while
predicting interaction. The method extracts phylogenetic profiles of
proteins by using their sequence information. Combining the phylogenetic
profiles of two proteins by checking existence of homologs
in different species and fitting this combined profile into a statistical
model, it is possible to make predictions about the interaction status
of two proteins.
For this purpose, we apply a collection of pattern recognition
techniques on the dataset of combined phylogenetic profiles of protein
pairs. Support Vector Machines, Feature Extraction using ReliefF,
Naive Bayes Classification, K-Nearest Neighborhood Classification,
Decision Trees, and Random Forest Classification are the methods
we applied for finding the classification method that best predicts
the interaction status of protein pairs. Random Forest Classification
outperformed all other methods with a prediction accuracy of 76.93%
Abstract: Random and natural textures classification is still
one of the biggest challenges in the field of image processing and
pattern recognition. In this paper, texture feature extraction using
Slant Hadamard Transform was studied and compared to other
signal processing-based texture classification schemes. A
parametric SHT was also introduced and employed for natural
textures feature extraction. We showed that a subtly modified
parametric SHT can outperform ordinary Walsh-Hadamard
transform and discrete cosine transform. Experiments were carried
out on a subset of Vistex random natural texture images using a
kNN classifier.
Abstract: Classification of electroencephalogram (EEG) signals
extracted during mental tasks is a technique that is actively pursued
for Brain Computer Interfaces (BCI) designs. In this paper, we
compared the classification performances of univariateautoregressive
(AR) and multivariate autoregressive (MAR) models
for representing EEG signals that were extracted during different
mental tasks. Multilayer Perceptron (MLP) neural network (NN)
trained by the backpropagation (BP) algorithm was used to classify
these features into the different categories representing the mental
tasks. Classification performances were also compared across
different mental task combinations and 2 sets of hidden units (HU): 2
to 10 HU in steps of 2 and 20 to 100 HU in steps of 20. Five different
mental tasks from 4 subjects were used in the experimental study and
combinations of 2 different mental tasks were studied for each
subject. Three different feature extraction methods with 6th order
were used to extract features from these EEG signals: AR
coefficients computed with Burg-s algorithm (ARBG), AR
coefficients computed with stepwise least square algorithm (ARLS)
and MAR coefficients computed with stepwise least square
algorithm. The best results were obtained with 20 to 100 HU using
ARBG. It is concluded that i) it is important to choose the suitable
mental tasks for different individuals for a successful BCI design, ii)
higher HU are more suitable and iii) ARBG is the most suitable
feature extraction method.
Abstract: A typical definition of the Computer Aided Diagnosis
(CAD), found in literature, can be: A diagnosis made by a radiologist
using the output of a computerized scheme for automated image
analysis as a diagnostic aid. Often it is possible to find the expression
Computer Aided Detection (CAD or CADe): this definition
emphasizes the intent of CAD to support rather than substitute the
human observer in the analysis of radiographic images. In this article
we will illustrate the application of CAD systems and the aim of
these definitions.
Commercially available CAD systems use computerized
algorithms for identifying suspicious regions of interest. In this paper
are described the general CAD systems as an expert system
constituted of the following components: segmentation / detection,
feature extraction, and classification / decision making.
As example, in this work is shown the realization of a Computer-
Aided Detection system that is able to assist the radiologist in
identifying types of mammary tumor lesions. Furthermore this
prototype of station uses a GRID configuration to work on a large
distributed database of digitized mammographic images.
Abstract: The segmentation of mouth and lips is a fundamental
problem in facial image analyisis. In this paper we propose a method
for lip segmentation based on rg-color histogram. Statistical analysis
shows, using the rg-color-space is optimal for this purpose of a pure
color based segmentation. Initially a rough adaptive threshold selects
a histogram region, that assures that all pixels in that region are
skin pixels. Based on that pixels we build a gaussian model which
represents the skin pixels distribution and is utilized to obtain a
refined, optimal threshold. We are not incorporating shape or edge
information. In experiments we show the performance of our lip pixel
segmentation method compared to the ground truth of our dataset and
a conventional watershed algorithm.
Abstract: Hidden Markov Model (HMM) is a stochastic method
which has been used in various signal processing and character
recognition. This study proposes to use HMM to recognize Javanese
characters from a number of different handwritings, whereby HMM
is used to optimize the number of state and feature extraction. An
85.7 % accuracy is obtained as the best result in 16-stated vertical
model using pure HMM. This initial result is satisfactory for
prompting further research.
Abstract: Automatic detection of syllable repetition is one of the
important parameter in assessing the stuttered speech objectively.
The existing method which uses artificial neural network (ANN)
requires high levels of agreement as prerequisite before attempting to
train and test ANNs to separate fluent and nonfluent. We propose
automatic detection method for syllable repetition in read speech for
objective assessment of stuttered disfluencies which uses a novel
approach and has four stages comprising of segmentation, feature
extraction, score matching and decision logic. Feature extraction is
implemented using well know Mel frequency Cepstra coefficient
(MFCC). Score matching is done using Dynamic Time Warping
(DTW) between the syllables. The Decision logic is implemented by
Perceptron based on the score given by score matching. Although
many methods are available for segmentation, in this paper it is done
manually. Here the assessment by human judges on the read speech
of 10 adults who stutter are described using corresponding method
and the result was 83%.
Abstract: Keystroke authentication is a new access control system
to identify legitimate users via their typing behavior. In this paper,
machine learning techniques are adapted for keystroke authentication.
Seven learning methods are used to build models to differentiate user
keystroke patterns. The selected classification methods are Decision
Tree, Naive Bayesian, Instance Based Learning, Decision Table, One
Rule, Random Tree and K-star. Among these methods, three of them
are studied in more details. The results show that machine learning
is a feasible alternative for keystroke authentication. Compared to
the conventional Nearest Neighbour method in the recent research,
learning methods especially Decision Tree can be more accurate. In
addition, the experiment results reveal that 3-Grams is more accurate
than 2-Grams and 4-Grams for feature extraction. Also, combination
of attributes tend to result higher accuracy.
Abstract: A state of the art Speaker Identification (SI) system
requires a robust feature extraction unit followed by a speaker
modeling scheme for generalized representation of these features.
Over the years, Mel-Frequency Cepstral Coefficients (MFCC)
modeled on the human auditory system has been used as a standard
acoustic feature set for speech related applications. On a recent
contribution by authors, it has been shown that the Inverted Mel-
Frequency Cepstral Coefficients (IMFCC) is useful feature set for
SI, which contains complementary information present in high
frequency region. This paper introduces the Gaussian shaped filter
(GF) while calculating MFCC and IMFCC in place of typical
triangular shaped bins. The objective is to introduce a higher
amount of correlation between subband outputs. The performances
of both MFCC & IMFCC improve with GF over conventional
triangular filter (TF) based implementation, individually as well as
in combination. With GMM as speaker modeling paradigm, the
performances of proposed GF based MFCC and IMFCC in
individual and fused mode have been verified in two standard
databases YOHO, (Microphone Speech) and POLYCOST
(Telephone Speech) each of which has more than 130 speakers.
Abstract: In this paper, we propose a supervised method for
color image classification based on a multilevel sigmoidal neural
network (MSNN) model. In this method, images are classified into
five categories, i.e., “Car", “Building", “Mountain", “Farm" and
“Coast". This classification is performed without any segmentation
processes. To verify the learning capabilities of the proposed method,
we compare our MSNN model with the traditional Sigmoidal Neural
Network (SNN) model. Results of comparison have shown that the
MSNN model performs better than the traditional SNN model in the
context of training run time and classification rate. Both color
moments and multi-level wavelets decomposition technique are used
to extract features from images. The proposed method has been
tested on a variety of real and synthetic images.
Abstract: In this paper a novel approach for generalized image
retrieval based on semantic contents is presented. A combination of
three feature extraction methods namely color, texture, and edge
histogram descriptor. There is a provision to add new features in
future for better retrieval efficiency. Any combination of these
methods, which is more appropriate for the application, can be used
for retrieval. This is provided through User Interface (UI) in the
form of relevance feedback. The image properties analyzed in this
work are by using computer vision and image processing algorithms.
For color the histogram of images are computed, for texture cooccurrence
matrix based entropy, energy, etc, are calculated and for
edge density it is Edge Histogram Descriptor (EHD) that is found.
For retrieval of images, a novel idea is developed based on greedy
strategy to reduce the computational complexity. The entire system
was developed using AForge.Imaging (an open source product),
MATLAB .NET Builder, C#, and Oracle 10g. The system was tested
with Coral Image database containing 1000 natural images and
achieved better results.
Abstract: In this paper the problem of face recognition under variable illumination conditions is considered. Most of the works in the literature exhibit good performance under strictly controlled acquisition conditions, but the performance drastically drop when changes in pose and illumination occur, so that recently number of approaches have been proposed to deal with such variability. The aim of this work is to introduce an efficient local appearance feature extraction method based steerable pyramid (SP) for face recognition. Local information is extracted from SP sub-bands using LBP(Local binary Pattern). The underlying statistics allow us to reduce the required amount of data to be stored. The experiments carried out on different face databases confirm the effectiveness of the proposed approach.
Abstract: In this paper, we propose a practical digital music matching system that is robust to variation in sound qualities. The proposed system is subdivided into two parts: client and server. The client part consists of the input, preprocessing and feature extraction modules. The preprocessing module, including the music onset module, revises the value gap occurring on the time axis between identical songs of different formats. The proposed method uses delta-grouped Mel frequency cepstral coefficients (MFCCs) to extract music features that are robust to changes in sound quality. According to the number of sound quality formats (SQFs) used, a music server is constructed with a feature database (FD) that contains different sub feature databases (SFDs). When the proposed system receives a music file, the selection module selects an appropriate SFD from a feature database; the selected SFD is subsequently used by the matching module. In this study, we used 3,000 queries for matching experiments in three cases with different FDs. In each case, we used 1,000 queries constructed by mixing 8 SQFs and 125 songs. The success rate of music matching improved from 88.6% when using single a single SFD to 93.2% when using quadruple SFDs. By this experiment, we proved that the proposed method is robust to various sound qualities.
Abstract: In this work, we present an automatic vehicle detection
system for airborne videos using combined features. We propose a
pixel-wise classification method for vehicle detection using Dynamic
Bayesian Networks. In spite of performing pixel-wise classification,
relations among neighboring pixels in a region are preserved in the
feature extraction process. The main novelty of the detection scheme is
that the extracted combined features comprise not only pixel-level
information but also region-level information. Afterwards, tracking is
performed on the detected vehicles. Tracking is performed using
efficient Kalman filter with dynamic particle sampling. Experiments
were conducted on a wide variety of airborne videos. We do not
assume prior information of camera heights, orientation, and target
object sizes in the proposed framework. The results demonstrate
flexibility and good generalization abilities of the proposed method on
a challenging dataset.
Abstract: Texture classification is an important image processing
task with a broad application range. Many different techniques for
texture classification have been explored. Using sparse approximation
as a feature extraction method for texture classification is a relatively
new approach, and Skretting et al. recently presented the Frame
Texture Classification Method (FTCM), showing very good results on
classical texture images. As an extension of that work the FTCM is
here tested on a real world application as detection of abnormalities
in mammograms. Some extensions to the original FTCM that are
useful in some applications are implemented; two different smoothing
techniques and a vector augmentation technique. Both detection of
microcalcifications (as a primary detection technique and as a last
stage of a detection scheme), and soft tissue lesions in mammograms
are explored. All the results are interesting, and especially the results
using FTCM on regions of interest as the last stage in a detection
scheme for microcalcifications are promising.
Abstract: Heart sound is an acoustic signal and many techniques
used nowadays for human recognition tasks borrow speech recognition
techniques. One popular choice for feature extraction of accoustic
signals is the Mel Frequency Cepstral Coefficients (MFCC) which
maps the signal onto a non-linear Mel-Scale that mimics the human
hearing. However the Mel-Scale is almost linear in the frequency
region of heart sounds and thus should produce similar results with
the standard cepstral coefficients (CC). In this paper, MFCC is
investigated to see if it produces superior results for PCG based
human identification system compared to CC. Results show that the
MFCC system is still superior to CC despite linear filter-banks in
the lower frequency range, giving up to 95% correct recognition rate
for MFCC and 90% for CC. Further experiments show that the high
recognition rate is due to the implementation of filter-banks and not
from Mel-Scaling.
Abstract: Over the past decades, automatic face recognition has become a highly active research area, mainly due to the countless application possibilities in both the private as well as the public sector. Numerous algorithms have been proposed in the literature to cope with the problem of face recognition, nevertheless, a group of methods commonly referred to as appearance based have emerged as the dominant solution to the face recognition problem. Many comparative studies concerned with the performance of appearance based methods have already been presented in the literature, not rarely with inconclusive and often with contradictory results. No consent has been reached within the scientific community regarding the relative ranking of the efficiency of appearance based methods for the face recognition task, let alone regarding their susceptibility to appearance changes induced by various environmental factors. To tackle these open issues, this paper assess the performance of the three dominant appearance based methods: principal component analysis, linear discriminant analysis and independent component analysis, and compares them on equal footing (i.e., with the same preprocessing procedure, with optimized parameters for the best possible performance, etc.) in face verification experiments on the publicly available XM2VTS database. In addition to the comparative analysis on the XM2VTS database, ten degraded versions of the database are also employed in the experiments to evaluate the susceptibility of the appearance based methods on various image degradations which can occur in "real-life" operating conditions. Our experimental results suggest that linear discriminant analysis ensures the most consistent verification rates across the tested databases.
Abstract: In this paper we address the issue of classifying the fluorescent intensity of a sample in Indirect Immuno-Fluorescence (IIF). Since IIF is a subjective, semi-quantitative test in its very nature, we discuss a strategy to reliably label the image data set by using the diagnoses performed by different physicians. Then, we discuss image pre-processing, feature extraction and selection. Finally, we propose two ANN-based classifiers that can separate intrinsically dubious samples and whose error tolerance can be flexibly set. Measured performance shows error rates less than 1%, which candidates the method to be used in daily medical practice either to perform pre-selection of cases to be examined, or to act as a second reader.
Abstract: We introduce an algorithm based on the
morphological shared-weight neural network. Being nonlinear and
translation-invariant, the MSNN can be used to create better
generalization during face recognition. Feature extraction is
performed on grayscale images using hit-miss transforms that are
independent of gray-level shifts. The output is then learned by
interacting with the classification process. The feature extraction and
classification networks are trained together, allowing the MSNN to
simultaneously learn feature extraction and classification for a face.
For evaluation, we test for robustness under variations in gray levels
and noise while varying the network-s configuration to optimize
recognition efficiency and processing time. Results show that the
MSNN performs better for grayscale image pattern classification
than ordinary neural networks.
Abstract: Detecting protein-protein interactions is a central problem in computational biology and aberrant such interactions may have implicated in a number of neurological disorders. As a result, the prediction of protein-protein interactions has recently received considerable attention from biologist around the globe. Computational tools that are capable of effectively identifying protein-protein interactions are much needed. In this paper, we propose a method to detect protein-protein interaction based on substring similarity measure. Two protein sequences may interact by the mean of the similarities of the substrings they contain. When applied on the currently available protein-protein interaction data for the yeast Saccharomyces cerevisiae, the proposed method delivered reasonable improvement over the existing ones.