Search

In this paper, we propose a deep convolutional neural network for camera based wildfire detection. We train the neural network via transfer learning and use window based analysis strategy to increase the fire detection rate. To achieve computational efficiency, we calculate frequency response of the kernels in convolutional and dense layers and eliminate those filters with low energy impulse response. Moreover, to reduce the storage for edge devices, we compare the convolutional kernels in Fourier domain and discard similar filters using the cosine similarity measure in the frequency domain. We test the performance of the neural network with a variety of wildfire video clips and prune system performs as good as the regular network in daytime wild fire detection, and it also works well on some night wild fire video clips.

In this paper, we propose a deep convolutional neural network for camera based wildfire detection. We train the neural network via transfer learning and use window based analysis strategy to increase the fire detection rate. To achieve computational efficiency, we calculate frequency response of the kernels in convolutional and dense layers and eliminate those filters with low energy impulse response. Moreover, to reduce the storage for edge devices, we compare the convolutional kernels in Fourier domain and discard similar filters using the cosine similarity measure in the frequency domain. We test the performance of the neural network with a variety of wildfire video clips and the pruned system performs as good as the regular network in daytime wild fire detection, and it also works well on some night wild fire video clips.

In this paper, we introduce a video-based wildfire detection scheme based on a computationally efficient additive deep neural network, which we call AddNet. This AddNet is based on a multiplication-free vector operator, which performs only addition and sign manipulation operations. In this regard, we construct a dot product-like operation from the mf-operator and use it to define dense and convolutional feed-forwarding passes in AddNet. We train AddNet on images taken from forestry surveillance cameras. Our experiments show that AddNet can achieve a time-saving by 12.4% when compared to an equivalent regular convolutional neural network (CNN). Furthermore, the smoke recognition performance of AddNet is as good as regular CNNs and substantially better than binary-weight neural networks.

Date of Conference: 30 November - 3 December 2020 Conference name: 12th International Conference on Computational Collective Intelligence, ICCCI 2020 Real-time flame detection is crucial in video-based surveillance systems. We propose a vision-based method to detect flames using Deep Convolutional Generative Adversarial Neural Networks (DCGANs). Many existing supervised learning approaches using convolutional neural networks do not take temporal information into account and require a substantial amount of labeled data. To have a robust representation of sequences with and without flame, we propose a two-stage training of a DCGAN exploiting spatio-temporal flame evolution. Our training framework includes the regular training of a DCGAN with real spatio-temporal images, namely, temporal slice images, and noise vectors, and training the discriminator separately using the temporal flame images without the generator. Experimental results show that the proposed method effectively detects flame in video with negligible false-positive rates in real-time. A. Enis Çetin’s research is partially funded by NSF with grant number 1739396 and NVIDIA Corporation. B. Uğur Töreyin’s research is partially funded by TÜBİTAK 114E426, İTÜ BAP MGA-2017-40964 and MOA-2019-42321.

Chemical and infra-red sensors generate distinct responses under similar conditions because of sensor drift, noise or resolution errors. In this paper, we develop novel machine learning methods for detecting and identifying VOC and Ammonia vapor from time-series data obtained by uncalibrated chemical and infrared sensors. We process time-series sensor signals using deep neural networks (DNN). Three neural network algorithms are utilized for this purpose. Additive neural networks (termed AddNet) are based on a multiplication-devoid operator and consequently exhibit energy efficiency compared to regular neural networks. The second algorithm uses generative adversarial neural networks so as to expose the classifying neural network to more realistic data points in order to help the classifier network to deliver improved generalization. Finally, we use conventional convolutional neural networks as a baseline method. Our findings indicate that using raw time-series data obtained from uncalibrated sensors and processing them using deep-learning-based methods yield better results than using hand-crafted feature parameters.

The problem of detecting the presence of Social Anxiety Disorder (SAD) using Electroencephalography (EEG) for classification has seen limited study and is addressed with a new approach that seeks to exploit the knowledge of EEG sensor spatial configuration. Two classification models, one which ignores the configuration (model 1) and one that exploits it with different interpolation methods (model 2), are studied. Performance of these two models is examined for analyzing 34 EEG data channels each consisting of five frequency bands and further decomposed with a filter bank. The data are collected from 64 subjects consisting of healthy controls and patients with SAD. Validity of our hypothesis that model 2 will significantly outperform model 1 is borne out in the results, with accuracy $6$--$7\%$ higher for model 2 for each machine learning algorithm we investigated. Convolutional Neural Networks (CNN) were found to provide much better performance than SVM and kNNs.

Date of Conference: 12-17 May 2019 Conference Name: 44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 Early detection of wildfire smoke in real-time is essentially important in forest surveillance and monitoring systems. We propose a vision-based method to detect smoke using Deep Convolutional Generative Adversarial Neural Networks (DC-GANs). Many existing supervised learning approaches using convolutional neural networks require substantial amount of labeled data. In order to have a robust representation of sequences with and without smoke, we propose a two-stage training of a DCGAN. Our training framework includes, the regular training of a DCGAN with real images and noise vectors, and training the discriminator separately using the smoke images without the generator. Before training the networks, the temporal evolution of smoke is also integrated with a motion-based transformation of images as a pre-processing step. Experimental results show that the proposed method effectively detects the smoke images with negligible false positive rates in real-time. The Institute of Electrical and Electronics Engineers Signal Processing Society

This paper addresses the problem of detecting epileptic seizures experienced by a human subject during sleep. Commonly used solutions to this problem mostly rely on detecting motion due to seizures using contact-based sensors or video-based sensors. We seek a low-cost, low-power alternative that can sense motion without making direct contact with the subject and provides high detection accuracy. We investigate the use of Passive InfraRed (PIR) sensors to sense human body motion caused by epileptic seizures during sleep which makes the body shake and causes the PIR sensor to generate an oscillatory output signal. This signal can be distinguished from that of ordinary motions during sleep using analysis with machine learning algorithms. The supervised hidden Markov model algorithm (HMM) and a 1-D and 2-D convolutional neural network (ConvNet) are used to classify the data set of the PIR sensor output into the occurrence of epileptic seizures, ordinary motions, or absence of motion. The method was tested on the PIR signals captured at 1 m from 33 recruited healthy subjects who, after watching seizure videos, either moved their body on a bed to simulate a seizure, ordinary motion, or lay still. The HMM algorithm attained 97.03% accuracy, while 1D-ConvNet and 2D-ConvNet attained an accuracy of 96.97% and 98.98%, respectively. All simulated seizures were successfully detected, with errors occurring only in distinguishing between ordinary motion and no motion, thereby demonstrating the potential for using PIR sensors in the epileptic seizure detection.

Date of Conference: 15-20 April 2018 Contrast enhancement is an important preprocessing step for the analysis of images. The main aim of contrast enhancement techniques is to increase the visibility of the objects by modifying the spatial characteristics of the image. In this paper, phase transition based contrast enhancement framework is proposed to overcome the limitations of existing image enhancement techniques. The proposed image enhancement framework transforms the changes in image phase into the variations of magnitude to enhance the structural details of the image and to improve visibility. In addition, the concept of Selective Variation (SV) technique is introduced and enhancement parameters are optimized using SV. The experimental studies that were carried out on TID2008 dataset, show that the proposed scheme obtains promising results on contrast enhancement.

Date of Conference: 28 Aug.-2 Sept. 2017 Conference name: 25th European Signal Processing Conference (EUSIPCO), 2017 We introduce a time-frequency scattering method using hyperbolic tangent function for vessel sound classification. The sound data is wavelet transformed using a two channel filter-bank and filter-bank outputs are scattered using tanh function. A feature vector similar to mel-scale cepstrum is obtained after a wavelet packed transform-like structure approximating the mel-frequency scale. Feature vectors of vessel sounds are classified using a support vector machine (SVM). Experimental results are presented and the new feature extraction method produces better classification results than the ordinary Mel-Frequency Cepstral Coefficients (MFCC) vectors. © EURASIP 2017.

In this paper, Hematoxylin and Eosin (H&E) stained liver images are classified by using both Local Binary Patterns (LBP) and one dimensional SIFT (1-D SIFT) algorithm. In order to obtain more meaningful features from the LBP histogram, a new feature vector extraction process is implemented for 1-D SIFT algorithm. LBP histograms are extracted with different approaches and concatenated with color histograms of the images. It is experimentally shown that,with the proposed approach, it possible to classify the H&E stained liver images with the accuracy of 88 % .

In this article, a novel algorithm for denoising images corrupted by impulsive noise is presented. Impulsive noise generates pixels whose gray level values are not consistent with the neighboring pixels. The proposed denoising algorithm is a two-step procedure. In the first step, image denoising is formulated as a convex optimization problem, whose constraints are defined as limitations on local variations between neighboring pixels. We use Projections onto the Epigraph Set of the TV function (PES-TV) to solve this problem. Unlike other approaches in the literature, the PES-TV method does not require any prior information about the noise variance. It is only capable of utilizing local relations among pixels and does not fully take advantage of correlations between spatially distant areas of an image with similar appearance. In the second step, a Wiener filtering approach is cascaded to the PES-TV-based method to take advantage of global correlations in an image. In this step, the image is first divided into blocks and those with similar content are jointly denoised using a 3D Wiener filter. The denoising performance of the proposed two-step method was compared against three state-of-the-art denoising methods under various impulsive noise models.

In this lecture note, we describe a wavelet domain denoising method consisting of making orthogonal projections of wavelet (subbands) signals of the noisy signal onto an upside down pyramid-shaped region in a multidimensional space. Each horizontal slice of the upside down pyramid is a diamond shaped region and it is called an -ball. The upside down pyramid is called the epigraph set of the -norm cost function. We show that the method leads to soft-thresholding as in standard wavelet denoising methods. Orthogonal projection operations automatically determine the soft-threshold values of the wavelet signals. © 2015 IEEE.

Date of Conference: 11-16 February 2017 Conference Name: SPIE Medical Imaging, 2017 The analysis and interpretation of histopathological samples and images is an important discipline in the diagnosis of various diseases, especially cancer. An important factor in prognosis and treatment with the aim of a precision medicine is the determination of so-called cancer stem cells (CSC) which are known for their resistance to chemotherapeutic treatment and involvement in tumor recurrence. Using immunohistochemistry with CSC markers like CD13, CD133 and others is one way to identify CSC. In our work we aim at identifying CSC presence on ubiquitous Hematoxilyn and Eosin (HE) staining as an inexpensive tool for routine histopathology based on their distinct morphological features. We present initial results of a new method based on color deconvolution (CD) and convolutional neural networks (CNN). This method performs favorably (accuracy 0.936) in comparison with a state-of-the-art method based on 1DSIFT and eigen-analysis feature sets evaluated on the same image database. We also show that accuracy of the CNN is improved by the CD pre-processing.

Follicular lymphoma (FL) is a group of malignancies of lymphocyte origin that arise from lymph nodes, spleen, and bone marrow in the lymphatic system. It is the second most common non-Hodgkins lymphoma. Characteristic of FL is the presence of follicle center B cells consisting of centrocytes and centroblasts. Typically, FL images are graded by an expert manually counting the centroblasts in an image. This is time consuming. In this paper, we present a novel multi-scale directional filtering scheme and utilize it to classify FL images into different grades. Instead of counting the centroblasts individually, we classify the texture formed by centroblasts. We apply our multi-scale directional filtering scheme in two scales and along eight orientations, and use the mean and the standard deviation of each filter output as feature parameters. For classification, we use support vector machines with the radial basis function kernel. We map the features into two dimensions using linear discriminant analysis prior to classification. Experimental results are presented.

This paper proposes a vessel recognition and classification system based on vessel acoustic signatures. Teager Energy Operator (TEO) based Mel Frequency Cepstral Coefficients (MFCC) are used for the first time in Underwater Acoustic Signal Recognition (UASR) to identify platforms the acoustic noise they generate. TEO based MFCC (TEO-MFCC), being more robust in noisy conditions than conventional MFCC, provides a better estimation platform energy. Conventionally, acoustic noise is recognized by sonar oper-ators who listen to audio signals received by ship sonars. The aim of this work is to replace this conventional human-based recognition system with a TEO-MFCC features-based classification system. TEO is applied to short-time Fourier transform (STFT) of acoustic signal frames and Mel-scale filter bank is used to obtain Mel Teager-energy spectrum. The feature vector is constructed by discrete cosine transform (DCT) of logarithmic Mel Teager-energy spectrum. Obtained spectrum is transformed into cepstral coefficients that are labeled as TEO-MFCC. This analysis and implementation are carried out with datasets of 24 different noise recordings that belong to 10 separate classes of vessels. These datasets are partially provided by National Park Service (NPS). Artificial Neural Networks (ANN) are used as a classification method. Experimental results demonstrate that TEO-MFCC achieves 99.5% accuracy in classification of vessel noises. © 2016 IEEE.

Assessment of visual quality plays a crucial role in modeling, implementation, and optimization of image- and video-processing applications. The image quality assessment (IQA) techniques basically extract features from the images to generate objective scores. Feature-based IQA methods generally consist of two complementary phases: (1) feature extraction and (2) feature pooling. For feature extraction in the IQA framework, various algorithms have been used and recently, the two-dimensional (2-D) mel-cepstrum (2-DMC) feature extraction scheme has provided promising results in a feature-based IQA framework. However, the 2-DMC feature extraction scheme completely loses image-phase information that may contain high-frequency characteristics and important structural components of the image. In this work, “2-D complex mel-cepstrum” is proposed for feature extraction in an IQA framework. The method tries to integrate Fourier transform phase information into the 2-DMC, which was shown to be an efficient feature extraction scheme for assessment of image quality. Support vector regression is used for feature pooling that provides mapping between the proposed features and the subjective scores. Experimental results show that the proposed technique obtains promising results for the IQA problem by making use of the image-phase information.

Gender classification, a two-class problem (male or female), has been the subject of extensive research recently and gained a lot of attention due to its varied set of applications. The proposed work relies on individual facial features to train a convolutional neural network (CNN) for gender classification. In contrast with previously reported results that assume the facial features are independent, we consider the facial features as correlated features by training a single CNN that jointly learns from all facial features. In terms of accuracy, our results either outperform, or are on par with, other gender classification techniques applied to three different datasets namely specs on faces, groups, and face recognition technology. In terms of performance, the proposed CNN has significantly fewer parameters as compared with other techniques reported in the literature. Our learnable parameters are fewer than those required in techniques reported in recent work, which enables them to make the network less sensitive to over-fitting and easier to train than techniques that use different CNNs for each facial feature as reported in the literature.

Connectivity-Guided Adaptive Wavelet Transform based mesh compression framework is proposed. The transformation uses the connectivity information of the 3D model to exploit the inter-pixel correlations. Orthographic projection is used for converting the 3D mesh into a 2D image-like representation. The proposed conversion method does not change the connectivity among the vertices of the 3D model. There is a correlation between the pixels of the composed image due to the connectivity of the 3D mesh. The proposed wavelet transform uses an adaptive predictor that exploits the connectivity information of the 3D model. Known image compression tools cannot take advantage of the correlations between the samples. The wavelet transformed data is then encoded using a zero-tree wavelet based method. Since the encoder creates a hierarchical bitstream, the proposed technique is a progressive mesh compression technique. Experimental results show that the proposed method has a better rate distortion performance than MPEG-3DGC/MPEG-4 mesh coder. © 2009 Elsevier Inc. All rights reserved.

In this paper, a video based algorithm for fire and flame detection is developed. In addition to ordinary motion and color clues, flame flicker is distinguished from motion of flame colored moving objects using Markov models. Irregular nature of flame boundaries is detected by performing temporal wavelet analysis using Hidden Markov Models as well. Color variations in fire is detected by computing the spatial wavelet transform of moving fire-colored regions. Boundary of flames are represented in wavelet domain and irregular nature of the boundaries of fire regions is also used as an indication of the flame flicker. Decisions from sub-algorithms are linearly combined using an adaptive active fusion method. The main detection algorithm is composed of four sub-algorithms (i) detection of fire colored moving objects, (ii) temporal, and (iii) spatial wavelet analysis for flicker detection and (iv) contour analysis of fire colored region boundaries. Each algorithm yields a continuous decision value as a real number in the range [-1, 1] at every image frame of a video sequence. Decision values from sub-algorithms are fused using an adaptive algorithm in which weights are updated using the least mean square (LMS) method in the training (learning) stage.

Cataloged from PDF version of article. In this paper we present a video based text and equation editor for LaTeX. The system recognizes what is written onto paper and generates the LaTeX code. Text and equations are written on a regular paper using a board marker, and a USB camera attached to a computer is used to capture and record the pen-tip positions in each consecutive image frame. Characters and symbols are represented as separate finite state machines (FSMs). They are written in an isolated manner and they are recognized on-line using the FSMs. In the last step, LaTeX code corresponding to recognized characters and symbols is generated. (c) 2007 Elsevier Ltd. All rights reserved.

Date of Conference: 26-28 November 2015 Conference Name: 9th International Conference on Electrical and Electronics Engineering, ELECO 2015 Respiratory rate is an essential parameter in many practical applications such as patient and elderly people monitoring. In this paper, a novel contact-free system is introduced to detect the human breathing activity. The system, which consists of two pyro-electric infrared (PIR) sensors, is capable of estimating the respiratory rate and detecting the sleep apnea. Sensors' signals corresponding to the thoracic movements of a human being are sampled using a microprocessor and analyzed on a general-purpose computer. Sampled signals are processed using empirical mode decomposition (EMD) and a new average magnitude difference function (AMDF) is used to detect the periodicity and the period of the processed signals. The resulting period, by using the fact that breathing is almost a periodic activity, is monitored as the respiratory rate. The new AMDF provides a way to fuse the data from the multiple sensors and generate a more reliable estimation of the respiratory rate.

This paper presents a cancer stem cell detection method using region covariance and codifference method. It focuses on detection of Cancer Stem Cell (CSC) in microscopic images which are stained with CD13 marker. Features of CSC images are extracted by using both covariance method and its multiplication free version codifference method and these features are fed into a Support Vector Machine (SVM) for classification. Experimental results are presented.

A new multi-resolution super-pixel based algorithm is proposed to track cell size, count and motion in Mesenchymal Stem Cells (MSCs) images. Multi-resolution super-pixels are obtained by placing varying density seeds on the image. The density of the seeds are determined according to the local high frequency components of the MSCs image. In this way a multi-resolution super-pixels decomposition of the image is obtained. A second contribution of the paper is novel decision rule for merging similar neighboring super-pixels. One-dimensional version of the well known scale invariant feature transform (SIFT) is developed and applied to the histograms of the neighboring super-pixels to determine similar regions. The proposed algorithm is experimentally shown to be successful in segmenting and tracking cells in MSCs images.

In this article, a new kernel estimation method is introduced using the epigraph set of the l 1 -norm. The new method produces a high-resolution and cross-term free estimates for Cohen's Class of Time-frequency (TF) distributions. The kernel estimation process starts with an initial rough TF distribution. This initial estimate is orthogonally projected onto the epigraph set of the l 1 norm in TF domain. Epigraph set of the l 1 norm produces a sparse time-frequency distribution. Sparsity in TF domain leads to cross-term free TF distributions. Experimental results are presented and the TF distributions obtained with the estimated kernel are compared to those obtained with an optimized kernel.

Date of Conference: 1-2 Nov. 2014 Conference name: International Workshop on Computational Intelligence for Multimedia Understanding, IWCIM 2014 The cosine similarity measure is widely used in big data analysis to compare vectors. In this article a new set of vector similarity measures are proposed. New vector similarity measures are based on a multiplication-free operator which requires only additions and sign operations. A vector 'product' using the multiplication-free operator is also defined. The new vector product induces the ℓ1-norm. As a result, new cosine measure-like similarity measures are normalized by the ℓ1-norms of the vectors. They can be computed using the MapReduce framework. Simulation examples are presented. © 2014 IEEE.

A system for removing shell pieces from hazelnut kernels using impact vibration analysis was developed in which nuts are dropped onto a steel plate and the vibration signals are captured and analyzed. The mel-cepstral feature parameters, line spectral frequency values, and Fourier-domain Lebesgue features were extracted from the vibration signals. The best experimental results were obtained using the melcepstral feature parameters. The feature parameters were classified using a support vector machine (SVM), which was trained a priori using a manually classified dataset. An average recognition rate of 98.2% was achieved. An important feature of the method is that it is easily trainable, enabling it to be applicable to other nuts, including walnuts and pistachio nuts. In addition, the system can be implemented in real time. 2013 Elsevier B.V. All rights reserved

Microscopic images play an essential role in detection and diagnosis of numerous diseases. For example, analysis of histology images of the human tissue biopsies remains the most reliable way of diagnosis and grading cancer. On the other hand, there is significant interand intra-rater variability in the grading and diagnosis of cancer from histology slides by clinicians. Computational histopathology can assist the pathologists in making the grading and diagnosis reproducible, by providing useful quantitative measures from histology images of a patient suspected or diagnosed of having cancer [1]. Similarly, in cellular biology, live cell fluorescence imagingmade possible by the rapid advances of fluorescent microscopy opens new avenues for microscopic image processing. It is possible to give many similar examples. In this special issue, we bring together six recent research articles on microscopic image processing and analysis. The subject topics of papers cover a diverse set of topics related to microscopic images processing such as creating efficient representations, segmentation of diagnostically significant structures, and automated classification. Even if the targeted applications are diverse, the main goal of all the algorithms was similar; they describe automated analysis methods for

This work characterizes the dynamics of sleep spindles, observed in electroencephalogram (EEG) recorded from humans during sleep, using both time and frequency domain methods which depend on higher order statistics and spectra. The time domain method combines the use of second- and third-order correlations to reveal information on the stationarity of periodic spindle rhythms to detect transitions between multiple activities. The frequency domain method, based on normalized spectrum and bispectrum, describes frequency interactions associated with nonlinearities occurring in the observed EEG.

In this article, image histogram thresholding is carried out using the likelihood of a mixture of Gaussians. In the proposed approach, a probability density function (PDF) of the histogram is computed using Gaussian kernel density estimation in an iterative manner. The threshold is found by iteratively computing a mixture of Gaussians for the two clusters. This process is aborted when the current bin is assigned to a different cluster than its predecessor. The method does not envolve an exhaustive search. Visual examples of our segmentation versus Otsu's thresholding method are presented.

In this paper, a Compressive Sensing based feature extraction algorithm is proposed for flame detection using infrared cameras. First, bright and moving regions in videos are detected. Then the videos are divided into spatio-temporal blocks and spatial and temporal feature vectors are exctracted from these blocks. Compressive Sensing is used to exctract spatial feature vectors. Compressed measurements are obtained by multiplying the pixels in the block with the sensing matrix. A new method is also developed to generate the sensing matrix. A random vector generated according to standard Gaussian distribution is passed through a wavelet transform and the resulting matrix is used as the sensing matrix. Temporal features are obtained from the vector that is formed from the difference of mean intensity values of the frames in two neighboring blocks. Spatial feature vectors are classified using Adaboost. Temporal feature vectors are classified using hidden Markov models. To reduce the computational cost only moving and bright regions are classified and classification is performed at specified intervals instead of every frame.

Date of Conference: 26-31 May 2013 The common spatial pattern (CSP) method has large number of applications in brain machine interfaces (BMI) to extract features from the multichannel neural activity through a set of linear spatial projections. These spatial projections minimize the Rayleigh quotient (RQ) as the objective function, which is the variance ratio of the classes. The CSP method easily overfits the data when the number of training trials is not sufficiently large and it is sensitive to daily variation of multichannel electrode placement, which limits its applicability for everyday use in BMI systems. To overcome these problems, the amount of channels that is used in projections, should be limited to some adequate number. We introduce a spatially sparse projection (SSP) method that renders unconstrained minimization possible via a new objective function with an approximated ℓ1 penalty. We apply our new algorithm with a baseline regularization to the ECoG data involving finger movements to gain stability with respect to the number of sparse channels. © 2013 IEEE.

Cancer cell lines are widely used for research purposes in laboratories all over the world. Computer-assisted classification of cancer cells can alleviate the burden of manual labeling and help cancer research. In this paper, we present a novel computerized method for cancer cell line image classification. The aim is to automatically classify 14 different classes of cell lines including 7 classes of breast and 7 classes of liver cancer cells. Microscopic images containing irregular carcinoma cell patterns are represented by subwindows which correspond to foreground pixels. For each subwindow, a covariance descriptor utilizing the dual-tree complex wavelet transform (DT-[Formula: see text]WT) coefficients and several morphological attributes are computed. Directionally selective DT-[Formula: see text]WT feature parameters are preferred primarily because of their ability to characterize edges at multiple orientations which is the characteristic feature of carcinoma cell line images. A Support Vector Machine (SVM) classifier with radial basis function (RBF) kernel is employed for final classification. Over a dataset of 840 images, we achieve an accuracy above 98%, which outperforms the classical covariance-based methods. The proposed system can be used as a reliable decision maker for laboratory studies. Our tool provides an automated, time- and cost-efficient analysis of cancer cell morphology to classify different cancer cell lines using image-processing techniques, which can be used as an alternative to the costly short tandem repeat (STR) analysis. The data set used in this manuscript is available as supplementary material through http://signal.ee.bilkent.edu.tr/cancerCellLineClassificationSampleImages.html.

WOS: 000315154800074, Features extracted at salient points are used to construct a region covariance descriptor (RCD) for target tracking. In the classical approach, the RCD is computed by using the features at each pixel location, which increases the computational cost in many cases. This approach is redundant because image statistics do not change significantly between neighboring image pixels. Furthermore, this redundancy may decrease tracking accuracy while tracking large targets because statistics of flat regions dominate region covariance matrix. In the proposed approach, salient points are extracted via the Shi and Tomasi's minimum eigenvalue method over a Hessian matrix, and the RCD features extracted only at these salient points are used in target tracking. Experimental results indicate that the salient point RCD scheme provides comparable and even better tracking results compared to a classical RCD-based approach, scale-invariant feature transform, and speeded-up robust features-based trackers while providing a computationally more efficient structure., framework of TUBITAK 1007 Program [109A001], This study is supported by the project with number 109A001 in the framework of TUBITAK 1007 Program. The authors would like to thank A. Onur Karali for his efforts in video capture and helpful discussions and Dr. M. Alper Kutay for his support in this study.

The goal of Ambient Assisted Living (AAL) research is to improve the quality of life of the elderly and handicapped people and help them maintain an independent lifestyle with the use of sensors, signal processing and telecommunications infrastructure. Unusual human activity detection such as fall detection has important applications. In this paper, a fall detection algorithm for a low cost AAL system using vibration and passive infrared (PIR) sensors is proposed. The single-tree complex wavelet transform (ST-CWT) is used for feature extraction from vibration sensor signal. The proposed feature extraction scheme is compared to discrete Fourier transform and mel-frequency cepstrum coefficients based feature extraction methods. Vibration signal features are classified into ''fall'' and ''ordinary activity'' classes using Euclidean distance, Mahalanobis distance, and support vector machine (SVM) classifiers, and they are compared to each other. The PIR sensor is used for the detection of a moving person in a region of interest. The proposed system works in real-time on a standard personal computer.

Cataloged from PDF version of article. Covariance-matrix-based features were applied to the detection of popcorn infected by a fungus that causes a symptom called "blue-eye". This infection of popcorn kernels causes economic losses due to the kernels' poor appearance and the frequently disagreeable flavor of the popped kernels. Images of kernels were obtained to distinguish damaged from undamaged kernels using image-processing techniques. Features for distinguishing blue-eye-damaged from undamaged popcorn kernel images were extracted from covariance matrices computed using various image pixel properties. The covariance matrices were formed using different property vectors that consisted of the image coordinate values, their intensity values and the first and second derivatives of the vertical and horizontal directions of different color channels. Support Vector Machines (SVM) were used for classification purposes. An overall recognition rate of 96.5% was achieved using these covariance based features. Relatively low false positive values of 2.4% were obtained which is important to reduce economic loss due to healthy kernels being discarded as fungal damaged. The image processing method is not computationally expensive so that it could be implemented in real-time sorting systems to separate damaged popcorn or other grains that have textural differences. (C) 2012 Elsevier B.V. All rights reserved

In this paper, we present a novel method for classification of cancer cell line images using complex wavelet-based region covariance matrix descriptors. Microscopic images containing irregular carcinoma cell patterns are represented by randomly selected subwindows which possibly correspond to foreground pixels. For each subwindow, a new region descriptor utilizing the dual-tree complex wavelet transform coefficients as pixel features is computed. ℂWT as a feature extraction tool is preferred primarily because of its ability to characterize singularities at multiple orientations, which often arise in carcinoma cell lines, and approximate shift invariance property. We propose new dissimilarity measures between covariance matrices based on Kullback-Leibler (KL) divergence and L 2 -norm, which turn out to be as successful as the classical KL divergence, but with much less computational complexity. Experimental results demonstrate the effectiveness of the proposed image classification framework. The proposed algorithm outperforms the recently published eigenvalue-based Bayesian classification method.

Date of Conference: 21-23 March 2012 Conference name: 46th Annual Conference on Information Sciences and Systems (CISS), 2012 Cancer cell lines are widely used for research purposes in laboratories all over the world. In this paper, we present a novel method for cancer cell line image classification, which is very costly by conventional methods. The aim is to automatically classify 14 different classes of cell lines including 7 classes of breast and 7 classes of liver cancer cells. Microscopic images containing irregular carcinoma cell patterns are represented by randomly selected subwindows which possibly correspond to foreground pixels. For each subwindow, a correlation descriptor utilizing the fractional unbalanced wavelet transform coefficients and several morphological attributes as pixel features is computed. Directionally selective textural features are preferred primarily because of their ability to characterize singularities at multiple orientations, which often arise in carcinoma cell lines. A Support Vector Machine (SVM) classifier with Radial Basis Function (RBF) kernel is employed for final classification. Over a dataset of 280 images, we achieved an accuracy of 88.2%, which outperforms the classical correlation based methods. © 2012 IEEE.

Date of Conference: 23-26 Sept. 2012 Common Spatial Pattern algorithm (CSP) is widely used in Brain Machine Interface (BMI) technology to extract features from dense electrode recordings by using their weighted linear combination. However, the CSP algorithm, is sensitive to variations in channel placement and can easily overfit to the data when the number of training trials is insufficient. Construction of sparse spatial projections where a small subset of channels is used in feature extraction, can increase the stability and generalization capability of the CSP method. The existing 0 norm based sub-optimal greedy channel reduction methods are either too complex such as Backward Elimination (BE) which provided best classification accuracies or have lower accuracy rates such as Recursive Weight Elimination (RWE) and Forward Selection (FS) with reduced complexity. In this paper, we apply the Oscillating Search (OS) method which fuses all these greedy search techniques to sparsify the CSP filters. We applied this new technique on EEG dataset IVa of BCI competition III. Our results indicate that the OS method provides the lowest classification error rates with low cardinality levels where the complexity of the OS is around 20 times lower than the BE. © 2012 IEEE.

Cataloged from PDF version of article. This paper proposes a video-based fire detection system which uses color, spatial and temporal information. The system divides the video into spatio-temporal blocks and uses covariance-based features extracted from these blocks to detect fire. Feature vectors take advantage of both the spatial and the temporal characteristics of flame-colored regions. The extracted features are trained and tested using a support vector machine (SVM) classifier. The system does not use a background subtraction method to segment moving regions and can be used, to some extent, with non-stationary cameras. The computationally efficient method can process 320 x 240 video frames at around 20 frames per second in an ordinary PC with a dual core 2.2 GHz processor. In addition, it is shown to outperform a previous method in terms of detection performance.

This book contains the contributions presented at the International Workshop on Computational Intelligence for Multimedia Understanding, organized by the ERCIM working group on Multimedia Understanding Through Semantics, Computation, and Learning (MUSCLE) as an open forum to discuss the most recent advances in this field. The workshop took place in Pisa, Italy, during December 13-15, 2011.

Date of Conference: 18-20 April 2012 Features extracted at salient points in the image are used to construct region covariance descriptor (RCD) for target tracking purposes. In the classical approach, the RCD is computed by using the features at each pixel location and thus, increases the computational cost in the scenarios where large targets are tracked. The approach in which the features at each pixel location are used, is redundant in cases where image statistics do not change significantly between neighboring pixels. Furthermore, this may decrease the tracking accuracy while tracking large targets which have background dominating structures. In the proposed approach, the salient points are extracted via the Shi and Tomasi's minimum eigenvalue method and a descriptor based target tracking structure is constructed based on the features extracted only at these salient points. Experimental results indicate that the proposed method provides comparable and in some cases even better tracking results compared to the classical method while providing a computationally more efficient structure. © 2012 IEEE.

Date of Conference: 30 Sept.-3 Oct. 2012 In this paper, a Content Based Image Retrieval (CBIR) algorithm employing relevance feedback is developed. After each round of user feedback Biased Discriminant Analysis (BDA) is utilized to find a transformation that best separates the positive samples from negative samples. The algorithm determines a sparse set of eigenvectors by L1 based optimization of the generalized eigenvalue problem arising in BDA for each feedback round. In this way, a transformation matrix is constructed using the sparse set of eigenvectors and a new feature space is formed by projecting the current features using the transformation matrix. Transformations developed using the sparse signal processing method provide better CBIR results and computational efficiency. Experimental results are presented. © 2012 IEEE.

Date of Conference: 22-27 May 2011 Video fire detection system which uses a spatio-temporal covariance matrix of video data is proposed. This system divides the video into spatio-temporal blocks and computes covariance features extracted from these blocks to detect fire. Feature vectors taking advantage of both the spatial and the temporal characteristics of flame colored regions are classified using an SVM classifier which is trained and tested using video data containing flames and flame colored objects. Experimental results are presented. © 2011 IEEE.

Date of Conference: 27 April-1 May 2011 A hybrid state detection algorithm is presented for the estimation of baseline and movement states which can be used to trigger a free paced neuroprostethic. The hybrid model was constructed by fusing a multiclass Support Vector Machine (SVM) with a Hidden Markov Model (HMM), where the internal hidden state observation probabilities were represented by the discriminative output of the SVM. The proposed method was applied to the multichannel Electrocorticogram (ECoG) recordings of BCI competition IV to identify the baseline and movement states while subjects were executing individual finger movements. The results are compared to regular Gaussian Mixture Model (GMM)-based HMM with the same number of states as SVM-based HMM structure. Our results indicate that the proposed hybrid state estimation method out-performs the standard HMM-based solution in all subjects studied with higher latency. The average latency of the hybrid decoder was approximately 290ms. © 2011 IEEE.

In this article, an image feature extraction method based on two-dimensional (2D) Mellin cepstrum is introduced. The concept of one-dimensional (1D) mel-cepstrum that is widely used in speech recognition is extended to two-dimensions using both the ordinary 2D Fourier transform and the Mellin transform. The resultant feature matrices are applied to two different classifiers such as common matrix approach and support vector machine to test the performance of the mel-cepstrum- and Mellin-cepstrum-based features. The AR face image database, ORL database, Yale database and FRGC database are used in experimental studies, which indicate that recognition rates obtained by the 2D mel-cepstrum-based method are superior to that obtained using 2D principal component analysis, 2D Fourier-Mellin transform and ordinary image matrix-based face recognition in both classifiers. Experimental results indicate that 2D cepstral analysis can also be used in other image feature extraction problems. © The Author 2010. Published by Oxford University Press on behalf of The British Computer Society.

Date of Conference: 30 Aug.-3 Sept. 2011 We tackle the problem of classifying multichannel electrocorticogram (ECoG) related to individual finger movements for a brain machine interface (BMI). For this particular aim we applied a recently developed hierarchical spatial projection framework of neural activity for feature extraction from ECoG. The algorithm extends the binary common spatial patterns algorithm to multiclass problem by constructing a redundant set of spatial projections that are tuned for paired and group-wise discrimination of finger movements. The groupings were constructed by merging the data of adjacent fingers and contrasting them to the rest, such as the first two fingers (thumb and index) vs. the others (middle, ring and little). We applied this framework to the BCI competition IV ECoG data recorded from three subjects. We observed that the maximum classification accuracy was obtained from the gamma frequency band (65200Hz). For this particular frequency range the average classification accuracy over three subjects was 86.3%. These results indicate that the redundant spatial projection framework can be used successfully in decoding finger movements from ECoG for BMI. © 2011 IEEE.

A multilead electrocardiography (ECG) data compression method is presented. First, a linear transform is applied to the standard ECG lead signals, which are highly correlated with each other. In this way a set of uncorrelated transform domain signals is obtained. Then, the resulting transform domain signals are compressed using various coding methods, including multirate signal processing and transform domain coding techniques. >

Date of Conference: 4-6 Oct. 2010 Most video processing applications require object tracking as it is the base operation for real-time implementations such as surveillance, monitoring and video compression. Therefore, accurate tracking of an object under varying scene conditions is crucial for robustness. It is well known that illumination variations on the observed scene and target are an obstacle against robust object tracking causing the tracker lose the target. In this paper, a 2D-cepstrum based approach is proposed to overcome this problem. Cepstral domain features extracted from the target region are introduced into the covariance tracking algorithm and it is experimentally observed that 2D-cepstrum analysis of the target object provides robustness to varying illumination conditions. Another contribution of the paper is the development of the co-difference matrix based object tracking instead of the recently introduced covariance matrix based method. ©2010 IEEE.