Ergun Erçelebi

Classification of EEG signals using neural network and logistic regression

Epileptic seizures are manifestations of epilepsy. Careful analyses of the electroencephalograph (EEG) records can provide valuable insight and improved understanding of the mechanisms causing epileptic disorders. The detection of epileptiform discharges in the EEG is an important component in the diagnosis of epilepsy. As EEG signals are non-stationary, the conventional method of frequency analysis is not highly successful in diagnostic classification. This paper deals with a novel method of analysis of EEG signals using wavelet transform and classification using artificial neural network (ANN) and logistic regression (LR). Wavelet transform is particularly effective for representing various aspects of non-stationary signals such as trends, discontinuities and repeated patterns where other signal processing approaches fail or are not as effective. Through wavelet decomposition of the EEG records, transient features are accurately captured and localized in both time and frequency context. In epileptic seizure classification we used lifting-based discrete wavelet transform (LBDWT) as a preprocessing method to increase the computational speed. The proposed algorithm reduces the computational load of those algorithms that were based on classical wavelet transform (CWT). In this study, we introduce two fundamentally different approaches for designing classification models (classifiers) the traditional statistical method based on logistic regression and the emerging computationally powerful techniques based on ANN. Logistic regression as well as multilayer perceptron neural network (MLPNN) based classifiers were developed and compared in relation to their accuracy in classification of EEG signals. In these methods we used LBDWT coefficients of EEG signals as an input to classification system with two discrete outputs: epileptic seizure or non-epileptic seizure. By identifying features in the signal we want to provide an automatic system that will support a physician in the diagnosing process. By applying LBDWT in connection with MLPNN, we obtained novel and reliable classifier architecture. The comparisons between the developed classifiers were primarily based on analysis of the receiver operating characteristic (ROC) curves as well as a number of scalar performance measures pertaining to the classification. The MLPNN based classifier outperformed the LR based counterpart. Within the same group, the MLPNN based classifier was more accurate than the LR based classifier.

Electrocardiogram signals de-noising using lifting-based discrete wavelet transform.

This paper introduces an effective technique for the denoising of electrocardiogram (ECG) signals corrupted by nonstationary noises. The technique is based on a second generation wavelet transform and level-dependent threshold estimator. Here, wavelet coefficients of ECG signals were obtained with lifting-based wavelet filters. A lifting scheme is used to construct second-generation wavelets and is an alternative and faster algorithm for a classical wavelet transform. The overall denoising performance of our proposed method is considered in relation to several measuring parameters, including types of wavelet filters (Haar, Daubechies 4 (DB4), Daubechies 6 (DB6), Filter(9-7), and Cubic B-splines), thresholding method, and decomposition depth. Three different kinds of noise were considered in this work: muscle artifact noise, electrode motion artifact noise, and white noise. Global performance is evaluated by means of the signal-to-noise ratio and visual inspection. Numerical results comparing the performance of the proposed method with that of nonlinear filtering techniques (median filter) are given. The results demonstrate consistently superior denoising performance of the proposed method over median filtering.

Second generation wavelet transform-based pitch period estimation and voiced/unvoiced decision for speech signals

Abstract Pitch detection is an important part of speech recognition and speech processing. In this paper, a pitch detection algorithm based on second generation wavelet transform was developed. The proposed algorithm reduces the computational load of those algorithms that were based on classical wavelet transform. The proposed pitch detection algorithm was tested for both real speech and synthetic speech signal. Some experiments were carried out under noisy environment condition to evaluate the accuracy and robustness of the proposed algorithm. Results showed that the proposed algorithm was robust to noise and provided accurate estimates of the pitch period for both low-pitched and high-pitched speakers. Moreover, different wavelet filters that were obtained using second generation wavelet transform were considered to see the effects of them on the proposed algorithm. It was noticed that Haar filter showed good performance as compared to the other wavelet filters.

Image enhancement via space-adaptive lifting scheme exploiting subband dependency

This paper proposes an image enhancement method based on space-adaptive, 2-D lifting scheme. In the space-adaptive update-first lifting scheme, the prediction stage is adapted to the signal structure point-by-point which results in a better signal representation and enhancement result. In this paper, a novel edge-sensitive adaptive prediction method is introduced in the 2-D lifting framework. The method adaptively chooses the best predictor among a set of predictors minimizing the prediction error. The proposed prediction method is sensitive to both even and odd indexed edge pixels in the 2-D lifting context. The bivariate shrinkage which assumes the dependence of the subband wavelet coefficients is used for subband image enhancement. As an objective quality measure, the peak signal-to-noise ratio test is applied to the results of the proposed image enhancement algorithm. Results of the proposed algorithm are compared with those of the VisuShrink, BayesShrink, and NorShrink. Experimental and objective quality test results prove the superior performance of the proposed image enhancement method.

Image enhancement via space-adaptive lifting scheme using spatial domain adaptive Wiener filter

This paper proposes an image enhancement method which is based on space-adaptive lifting scheme. The space-adaptive lifting scheme is designed with update-first strategy and the spatial domain adaptive Wiener filter is used for subband image enhancement. The space-adaptive lifting scheme is superior to the scale-adaptive one in that, the prediction stage is adapted to the signal structure pixel-by-pixel (not scale-by-scale) which results in a better signal representation and better image enhancement. A novel adaptive prediction method is proposed in the 2D lifting scheme. The method adaptively chooses the best prediction filter (predictor), among a set of predictors, based on minimizing the prediction error (or detail coefficient). The experimental results and objective quality test results prove the accomplishment of the proposed image enhancement method.

Classification of EEG for Epilepsy Diagnosis in Wavelet Domain Using Artifical Neural Network and Multi Linear Regression

In this study, classification methods were proposed for diagnosis of epilepsy in EEG signals using lifting based wavelet transform (LBWT) with artificial neural network (ANN) and multi linear regression (MLR). In classification of EEG signals, LBWT was used to increase computational speed in the extraction of the feature vectors. In comparison of LBWT with the classical wavelet transform, it was observed that LBWT decreased computational load as 50%. The coefficients in delta, theta, alpha, and beta bands that were obtained by LBWT were used as input signals of classifiers. ANN was trained as its output is logic 0 or logic 1 if EEG includes no epileptic seizure. The effects of different wavelet filters (Haar, Daubechies 4,6,8) on proposed methods were also observed. Proposed methods were compared from the point of accuracy, specify, and sensitivity. With this study, we aimed to provide an automatic decision support tool for neurologists treating potential epilepsy by defining features in EEG signals. We obtained a new and safe classifier using LBWT together with ANN

Robust multi bit and high quality audio watermarking using pseudo-random sequences

A robust multi bit and high quality audio watermarking technique in time domain is proposed in this paper. Watermarking is a technique used to label digital media by hiding copyright or other information into the underlying data. The watermark must be imperceptible and undetectable by the user and should be robust to various types of distortion. To enhance the robustness and taper-resistance of the embedded watermark, in this paper a multi bit technique is employed. Instead of embedding one bit into an audio frame, multiple bits can be embedded into each audio sub-frame. For attackers, since they do not know the parameters, this significantly reduces the possibility of unauthorized bit detection and removal of watermark. In real-time watermarking applications, robustness is not the only factor that plays an important role. The other factor that plays a very an important role is computational complexity. In general, audio file is transmitted in compressed form. Real-time watermark embedding must take into account this compressed form, because first decompressing the data, adding a watermark and then recompressing the data is computationally too demanding. In this paper, we propose robust and low complexity audio watermarking algorithm. To evaluate the performance of the proposed audio watermarking algorithm, subjective and objective quality tests including bit error rate (BER) and signal-to-noise ratio (SNR) were conducted. Compared to traditional one bit embedding algorithm, the proposed algorithm gives highly recovery rate after attack by commonly used audio data manipulations, such as low pass filtering, requantization, resampling, and MP3 compression.

A proportional derivative sliding mode control for an overactuated quadcopter

Traditional quadcopters suffer from their intrinsic underactuation, which prevents them from tracking arbitrary trajectories. In this study, a step-by-step mathematical modeling of a tilt rotor quadcopter, i.e. a quadcopter with all its four rotors are allowed to be tilted independently around their arms’ extension, is derived. The tilting mechanism converts the classical quadcopter to an overactuated flying vehicle that has full control over its states. The nonlinear dynamical model is derived based on the Newton–Euler formalization. A novel trajectory tracking control scheme is then proposed and developed. The proposed controller combines the proportional derivative linear controller with the nonlinear sliding mode controller. In order to reduce the chattering effect of the sliding mode controller, the discontinuous Signum switching function is replaced by a continuous sigmoidal function. The controller parameters are then tuned with the aid of genetic algorithm as an optimization tool. The genetic algorithm objective function is set so as to get the best step response characteristics. A simulation based analysis is used to proof the system and controller capability in following complex trajectories. Finally, the proposed controller robustness and effectiveness are analyzed. The simulation test results reveal the validity and feasibility of the proportional derivative sliding mode controller. The proposed controller also performed well in the face of modeling imprecision, sensor noise, and external disturbances.

A new teleconference system with a fast technique in HEVC coding

Tele-Communications play an important role nowadays. Globalization has led the requirement for traversed communication. Therefore a multimedia system takes an influential act in the comprehensive communication between the regions. Teleconferencing became an indispensable element in any business system. A teleconference offers the collaborators the ability to participate in a group consultation by managing a virtual convention while residing in geographically scattered areas. It also increases productivity, minimizes travel expenses and saves travel time. This paper presents a new Reliable Teleconference system that utilizes an improved HEVC (H.265) coding technology as an adequate approach to enhance the real-time video/IP technology system with improving the video quality and enhancing the compression efficiency corresponding with the previous codec (H.264).

A novel algorithm for blocking artifacts removal based on adaptive fuzzy filter in compressed images

In this study, a new adaptive post-filtering algorithm is proposed to remove observed blocking artifacts as a result of discrete cosine transform (DCT) based image and video compression standarts at low bit rates. With identification of blocking artifact strength, fuzzy filter is applied by adjusting filtering range and its parameters. Experimental results showed that, the proposed algorithm exhibits better detail preservation and artifact removal performance with lower computational cost as compared to other post-processing techniques. Accordingly, this can be used for the real time image and video applications without undesired artifacts.