In this work, we proposed a new gender detection algorithm based on pulmonary function test. The ... more In this work, we proposed a new gender detection algorithm based on pulmonary function test. The proposed method has three main stages. In first the stage, some features are extracted from pulmonary function test. In the second stage, the probabilistic models based on Gaussian Mixture Model (GMM) are trained using these features, and in the final stage, the gender of test person is detected based on likelihood ratio test. 1. Giris Biyometrik tanima, bireylere ait karakteristik ozelliklerin kullanilarak kisilerin taninmasini saglayan bir yontemdir. Bu yontem diger kart ve parola ile olusturulmus tanima sistemlerine gore daha guvenli ve etkin bir yontemdir. Son yillarda bu yontem ile gelistirilmis akilli uygulama sistemleri mevcuttur. Literaturde yuz, ses, konusma, parmak izi, el yazisi, retina gibi bircok biyometrik ozelligin tek basina ya da bir kacinin birlestirilmesi ile kisi tanima uzerine yapilmis calismalar mevcuttur (1, 2 ). Solunum fonksiyon testleri kisilerde solunum yollari...
2020 28th Signal Processing and Communications Applications Conference (SIU)
Almost all products on the market today have a unique code or ID associated with them. This speci... more Almost all products on the market today have a unique code or ID associated with them. This special identification is called a barcode. Barcodes have been the subject of extensive research in recent years due to the high demand for automation in various industrial environments. Fast and accurate reading of barcodes, where all details about the product used in many commercial applications can be learned, is very important. In this study, Mask R-CNN algorithm was used to determine the regions of the 1B barcodes in the image. In the Mask R-CNN, barcodes in the image have been detected, as well as the bounding box position of each barcode, as well as the pixel information corresponding to this class in the bounding box. Colored barcodes on various products taken at different ambient lights and at different angles were collected and a data set of 1114 images was prepared. Using this dataset, 74.41 % accuracy was achieved with Mask R-CNN.
2020 28th Signal Processing and Communications Applications Conference (SIU)
In this study, using the polysomnography data set, the classification of sleep stages was realize... more In this study, using the polysomnography data set, the classification of sleep stages was realized automatically with supervised learning method. In this study, sleep stage classification was carried out in three stages. In the first stage, the biomedical signal was divided into its independent components by Independent Component Analysis method. In the second stage, feature extraction was performed by using Mel Kepstrum Coefficient method. In the third stage, artificial neural networks were trained by using the extracted features and the sleep phases were estimated by using software architecture called Long ShortTerm Memory. As a result of the classification process performed in this way, the accuracy rate of the ten fold cross validation obtained for the binary classification (asleep / wake) was found to be 97,87%. For the five fold classification problem, the accuracy rate of the subject dependent algorithm and polysomnography data obtained using a single EEG channel was found to be 93,36% as a result of the ten-fold cross validation process.
2022 5th International Conference on Computing and Informatics (ICCI)
Convolutional neural networks (CNN), which have the advantage of automatically detecting the impo... more Convolutional neural networks (CNN), which have the advantage of automatically detecting the important features of the input data without any human interfere, are widely used in many applications such as face recognition, speech recognition, image classification and object detection. In real-time CNN applications, computation speed is very important as well as accuracy. However, in some applications with high computational complexity, available systems are insufficient to meet the high-speed performance demand at low power consumption. In this study, the design of the CNN accelerator hardware in FPGA is presented to meet the speed demand. In this design, CNN is considered as a streaming interface application. Thus, temporary storage amount and memory latency are reduced. Each layer is designed with maximum parallelism, taking advantage of the FPGA. Because fixed-point number representation has the advantage of low latency, it is preferred in design with negligible sacrifice of accuracy. Thus, forward propagation of a CNN can be executed at high speed in FPGA. In order to compare real-time performance, digit classification application is executed in this hardware designed in FPGA and ARM processor on the same chip. The real-time results show that the application in the hardware designed in the FPGA is 30x faster than the ARM processor.
The event related P300 potentials, positive waveforms in electroencephalography (EEG) signals, ar... more The event related P300 potentials, positive waveforms in electroencephalography (EEG) signals, are often utilized in brain computer interfaces (BCI). Many studies have been carried out to improve the performance of P300 speller systems either by developing signal processing algorithms and classifiers with different architectures or by designing new paradigms. In this study, a new paradigm is proposed for this purpose. The proposed paradigm combines two remarkable properties of being a 3D animation and utilizing column-only flashings as opposed to classical paradigms which are based on row-column flashings in 2D manner. The new paradigm is utilized in a traditional two-layer artificial neural networks model with a single output neuron, and numerous experiments are conducted to evaluate and compare the performance of the proposed paradigm with that of the classical approach. The experimental results, including statistical significance tests, are presented for single and multiple EEG e...
Erzincan Üniversitesi Fen Bilimleri Enstitüsü Dergisi, 2021
Satellite image analysis is a research area in which many research studies are carried out for ci... more Satellite image analysis is a research area in which many research studies are carried out for civil and military applications in the field of image processing. Satellite imagery has many applications including recognition, detection and classification of regions, buildings, roads, aircraft and other man-made objects. Among these, especially aircraft detection is strategically important for military applications, and forms the basis of this study. In the first phase of the study, a new dataset of aircrafts is created from Google Earth images to compensate the shortage of data set in this area. In the second stage, the detection of air vehicles was carried out using algorithms based on Convolutional Neural Network (CNN). Region-based Fully Convolutional Network (R-FCN), Single Shot Multi Box Detector (SSD) and Faster R-CNN methods are used for this process. The obtained accuracy rate for R-FCN, SSD and Faster R-CNN are 98.01%, 69.71% and 96.56%, respectively.
21st European Signal Processing Conference (EUSIPCO 2013), 2013
In this work, we have investigated the heart sound (HS) detection performance of Hidden Markov Mo... more In this work, we have investigated the heart sound (HS) detection performance of Hidden Markov Model (HMM) in respiratory sound. Respiratory sound is composed of heart sound and lung sound, and the main frequency components of these two sounds overlap with each other. To detect the locations of heart sound segments in such adverse condition accurately, the proposed method employs following steps. First, the Shannon entropy feature is extracted for robust representation of respiratory signal for different flow rates. Second, the probabilistic models are constructed by training HMM. Finally, the location of heart sound segments are efficiently estimated by the Viterbi decoding algorithm. The experimental results showed that the proposed heart sound detection method outperforms the three well-known heart sound detection methods in the literature. The average false negative rate (FNR) values for the proposed method are 5.4 ± 2.4 and 6.3 ± 1.3 for both low and medium respiratory flow rat...
2017 IEEE Global Conference on Signal and Information Processing (GlobalSIP), 2017
This paper examines filterbank parameters to extract discriminative cepstrum coefficient from EEG... more This paper examines filterbank parameters to extract discriminative cepstrum coefficient from EEG signals for sleep stage classification using well-known Support Vector Machine (SVM) algorithm. The proposed method has three main stages as feature extraction, training and classification. In feature extraction step, features are obtained using linear frequency cepstrum coefficients (LFCC) of EEG signals. Then SVM classifier is trained based on the extracted features. In the final step of classification, the class of test subject is estimated by using the trained model. Experimental results show that about an average of 95 percent correct classification rate is achievable for three classes, and this is better than the compared results available in the literature.
Abstract An efficient de-embedding air-line microwave method has been proposed for accurate relat... more Abstract An efficient de-embedding air-line microwave method has been proposed for accurate relative complex permittivity, e r = e r ′ − i e r ′ ′ , measurement of water-adulteration level within honey. It could be effectively applied to eliminate the errors arising from usage of imperfect calibration standards because it bypasses the requirement of these standards. Its accuracy is improved by utilizing the unitary and similarity properties of a passive two-port network, and then is compared with the accuracy of a calibration-dependent method present in the literature by using normalized root-mean-square-error (N-RMSE) values of e r ′ and e r ′ ′ of distilled water, in reference to the Debye value. From this comparison, it is observed that N-RMSE values calculated for e r ′ (and e r ′ ′ ) by using this calibration-dependent method and the (improved) proposed method are, respectively, around 0.1955 (0.1002) and 0.1962 (1.1067), indicating a good agreement between them. After validation the proposed de-embedding method using distilled water measurements, tested pure honey was adulterated with distilled water by different percentage values δ ranging from 1% to 10% in 1% increments. It is observed that the maximum distance between extracted e r ′ (or e r ′ ′ ) values of adulterated honey by the applied calibration-dependent method and the proposed method is less than 2%. Afterwards, an empirical formula was devised to fit e r ′ and e r ′ ′ values from measured e r of water-adulterated honey with various δ levels. It is noted that extracted e r ′ is much more better fitted than extracted e r ′ ′ , especially for δ ≤ 4 . Next, an optimization process is followed to evaluate the frequency for optimum prediction of adulteration levels using the empirical formula based on e r ′ or e r ′ ′ . It is noticed that optimized δ values using the empirical formula based on e r ′ (with an average prediction error of around 0.071 at 4.5 GHz) are superior to optimized δ values using the empirical formula based on e r ′ ′ (with an average prediction error of around 0.085 at 4.2 GHz) for prediction of previously known δ values. Sensitivity and uncertainty analyses were performed to assess and improve the accuracy of the proposed method.
Porous silicon (PSi) Fabry-Pérot (FP) cavities are the sensor types widely employed in sensing ap... more Porous silicon (PSi) Fabry-Pérot (FP) cavities are the sensor types widely employed in sensing applications of chemical, biological, or gas molecules. Prior to sensor operation, each fabricated empty (no-molecule) PSi FP cavity is characterized by their optical properties (refractive indices and thicknesses of each layer). For this characterization, a scanning electron microscope (SEM) image of a cut-sample from the batch FP cavity is measured to monitor the thickness variation of each layer. This technique is surely destructive and gives local information of only the broken cut-sample. In this Letter, we apply the full spectra fitting technique for nondestructive and accurate optical characterization of empty PSi FP cavities. To demonstrate the potential of this technique, we obtained the optimized thicknesses of each layer of two fabricated PSi FP cavities with resonating wavelengths of 542 nm and 1456 nm and compared them with the thicknesses obtained from their SEM images. From this analysis, we note that the proposed technique can be a good candidate for nondestructive characterization of empty FP cavities.
2015 23nd Signal Processing and Communications Applications Conference (SIU), 2015
In this study, it is aimed to produce microelectrodes which can be used in the detection of neuro... more In this study, it is aimed to produce microelectrodes which can be used in the detection of neurotransmitters that are related with brain disorders such as Parkinson, Epilepsy, and Schizophrenia and that exist in the central nervous system (CNS). A 4-channel, ceramic-based fabrication is performed towards this goal by using photolithographic methods. The time-current graphic response against the addition of H2O2 the produced microelectrode is analyzed in the calibration test. It is observed that the response is in stepwise form. In addition, limit of detection (LOD) of the produced microelectrodes and linearity values are shown to be within the desired ranges.
Medical & Biological Engineering & Computing, 2014
This work presents a detailed framework to detect the location of heart sound within the respirat... more This work presents a detailed framework to detect the location of heart sound within the respiratory sound based on temporal fuzzy c-means (TFCM) algorithm. In the proposed method, respiratory sound is first divided into frames and for each frame, the logarithmic energy features are calculated. Then, these features are used to classify the respiratory sound as heart sound (HS containing lung sound) and non-HS (only lung sound) by the TFCM algorithm. The TFCM is the modified version fuzzy c-means (FCM) algorithm. While the FCM algorithm uses only the local information about the current frame, the TFCM algorithm uses the temporal information from both the current and the neighboring frames in decision making. To measure the detection performance of the proposed method, several experiments have been conducted on a database of 24 healthy subjects. The experimental results show that the average false-negative rate values are 0.8 ± 1.1 and 1.5 ± 1.4 %, and the normalized area under detection error curves are 0.0145 and 0.0269 for the TFCM method in the low and medium respiratory flow rates, respectively. These average values are significantly lower than those obtained by FCM algorithm and by the other compared methods in the literature, which demonstrates the efficiency of the proposed TFCM algorithm. On the other hand, the average elapsed time of the TFCM for a data with length of 0.2 ± 0.05 s is 0.2 ± 0.05 s, which is slightly higher than that of the FCM and lower than those of the other compared methods.
This paper investigates the utility of a likelihood ratio test (LRT) combined with an efficient a... more This paper investigates the utility of a likelihood ratio test (LRT) combined with an efficient adaptation procedure for the purpose of detecting the heart sound (HS) with lung sound and the lung sound only (non-HS) segments in a respiratory signal. The proposed detection method has four main stages: feature extraction, training of the models, detection, and adaptation of the model parameter. In the first stage, the logarithmic energy features are extracted for each frame of respiratory sound. In the second stage, the probabilistic models for HS and non-HS segments are constructed by training Gaussian mixture models (GMMs) with an expectation maximization algorithm in a subject-independent manner, and then the HS and non-HS segments are detected by the results of the LRT based on the GMMs. In the adaptation stage, the subject-independent trained model parameter is modified online using the observed test data to fit the model parameter of the target subject. Experiments were performe...
We propose a novel method for identification of various gases using combined measurements of wave... more We propose a novel method for identification of various gases using combined measurements of wavelength shift (λ 0) and reflectivity difference (R). The method relies on plotting the measurement data on the R-λ 0 space and exploits in the identification process the effect of significant loss behavior of bulk silicon near or within the visible range (for lossless gas vapors). It has been validated by reflectivity measurements on a Fabry-Pérot cavity resonating around 745 nm for four different gas vapors whose reflective indices are close to each other and by two metrics for separability/identification analysis by the scattering matrix method. It is noted that the combined measurements of λ 0 and R result in better identification than that obtained by the measurement of λ 0 (or R) itself.
Three different techniques are applied for accurate constitutive parameters determination of isot... more Three different techniques are applied for accurate constitutive parameters determination of isotropic split-ring resonator (SRR) and SRR with a cut wire (Composite) metamaterial (MM) slabs. The first two techniques use explicit analytical calibrationdependent and calibration-invariant expressions while the third technique is based on Lorentz and Drude dispersion models. We have tested these techniques from simulated scattering (S-) parameters of two classic SRR and Composite MM slabs with various level of losses and different calibration plane factors. From the comparison, we conclude that whereas the extracted complex permittivity of both slabs by the analytical techniques produces unphysical results at resonance regions, that by the dispersion model eliminates this shortcoming and retrieves physically accurate constitutive parameters over the whole analyzed frequency region. We argue that incorrect retrieval of complex permittivity by analytical methods comes from spatial dispersion effects due to the discreteness of conducting elements within MM slabs which largely vary simulated S-parameters in the resonance regions where the slabs are highly spatially dispersive.
Heart sounds are the main obstacle in lung sound analysis. To tackle this obstacle, we propose a ... more Heart sounds are the main obstacle in lung sound analysis. To tackle this obstacle, we propose a diagnosis algorithm that uses singular spectrum analysis (SSA) and frequency features of heart and lung sounds. In particular, we introduce a frequency coefficient that shows the frequency difference between heart and lung sounds. The proposed algorithm is applied to a synthetic mixture of heart and lung sounds. The results show that heart sounds can be extracted successfully and localizations for the first and second heart sounds are remarkably performed. An error analysis of the localization results shows that the proposed algorithm has fewer errors compared to the SSA method, which is one of the most powerful methods in the localization of heart sounds. The presented algorithm is also applied in the cases of recorded respiratory sounds from the chest walls of five healthy subjects. The efficiency of the algorithm in extracting heart sounds from the recorded breathing sounds is verified with power spectral density evaluations and listening. Most studies have used only normal respiratory sounds, whereas we additionally use abnormal breathing sounds to validate the strength of our achievements.
Engineering Science and Technology, an International Journal, 2022
Drone detection and classification, important in military and civilian applications, are performe... more Drone detection and classification, important in military and civilian applications, are performed using different sensor signals. Proposed study handles this task using Radio Frequency (RF) signals utilizing basic machine learning methods. It is composed of two main stages as feature extraction succeeded by training/testing of the model. In feature extraction stage, valuable information for classification, contained in the RF signal, is obtained. For this purpose, spectral features, frequently used in speech processing applications, are employed. Specifically, Power Spectral Density (PSD), Mel-Frequency Cepstral Coefficients (MFCC) and Linear Frequency Cepstral Coefficients (LFCC) are adopted by adjusting filter bank margins and parameters for this task. In the second stage, a Support Vector Machine (SVM) classifier is first trained based on the obtained features and finally tested to measure its performance. All experimental studies are carried out using publicly available DroneRF dataset. This dataset contains 2-Class, 4-Class and 10-Class samples for drone existence vs. background (BG), drone types and drone operation modes, respectively. The best classification results are obtained using, PSD, MFCC and LFCC based features for 2-Class, MFCC and LFCC based features for 4-Class and LFCC based features for 10-Class. Accuracy rates for 2-Class, 4-Class and 10-Class are 100%, 98.67% and 95.15%, respectively. These results show that the proposed method outperforms the results given in the literature for DroneRF dataset.
Engineering Science and Technology, an International Journal, 2022
Drone detection and classification, important in military and civilian applications, are performe... more Drone detection and classification, important in military and civilian applications, are performed using different sensor signals. Proposed study handles this task using Radio Frequency (RF) signals utilizing basic machine learning methods. It is composed of two main stages as feature extraction succeeded by training/testing of the model. In feature extraction stage, valuable information for classification, contained in the RF signal, is obtained. For this purpose, spectral features, frequently used in speech processing applications, are employed. Specifically, Power Spectral Density (PSD), Mel-Frequency Cepstral Coefficients (MFCC) and Linear Frequency Cepstral Coefficients (LFCC) are adopted by adjusting filter bank margins and parameters for this task. In the second stage, a Support Vector Machine (SVM) classifier is first trained based on the obtained features and finally tested to measure its performance. All experimental studies are carried out using publicly available DroneRF dataset. This dataset contains 2-Class, 4-Class and 10-Class samples for drone existence vs. background (BG), drone types and drone operation modes, respectively. The best classification results are obtained using, PSD, MFCC and LFCC based features for 2-Class, MFCC and LFCC based features for 4-Class and LFCC based features for 10-Class. Accuracy rates for 2-Class, 4-Class and 10-Class are 100%, 98.67% and 95.15%, respectively. These results show that the proposed method outperforms the results given in the literature for DroneRF dataset.
In this work, we proposed a new gender detection algorithm based on pulmonary function test. The ... more In this work, we proposed a new gender detection algorithm based on pulmonary function test. The proposed method has three main stages. In first the stage, some features are extracted from pulmonary function test. In the second stage, the probabilistic models based on Gaussian Mixture Model (GMM) are trained using these features, and in the final stage, the gender of test person is detected based on likelihood ratio test. 1. Giris Biyometrik tanima, bireylere ait karakteristik ozelliklerin kullanilarak kisilerin taninmasini saglayan bir yontemdir. Bu yontem diger kart ve parola ile olusturulmus tanima sistemlerine gore daha guvenli ve etkin bir yontemdir. Son yillarda bu yontem ile gelistirilmis akilli uygulama sistemleri mevcuttur. Literaturde yuz, ses, konusma, parmak izi, el yazisi, retina gibi bircok biyometrik ozelligin tek basina ya da bir kacinin birlestirilmesi ile kisi tanima uzerine yapilmis calismalar mevcuttur (1, 2 ). Solunum fonksiyon testleri kisilerde solunum yollari...
2020 28th Signal Processing and Communications Applications Conference (SIU)
Almost all products on the market today have a unique code or ID associated with them. This speci... more Almost all products on the market today have a unique code or ID associated with them. This special identification is called a barcode. Barcodes have been the subject of extensive research in recent years due to the high demand for automation in various industrial environments. Fast and accurate reading of barcodes, where all details about the product used in many commercial applications can be learned, is very important. In this study, Mask R-CNN algorithm was used to determine the regions of the 1B barcodes in the image. In the Mask R-CNN, barcodes in the image have been detected, as well as the bounding box position of each barcode, as well as the pixel information corresponding to this class in the bounding box. Colored barcodes on various products taken at different ambient lights and at different angles were collected and a data set of 1114 images was prepared. Using this dataset, 74.41 % accuracy was achieved with Mask R-CNN.
2020 28th Signal Processing and Communications Applications Conference (SIU)
In this study, using the polysomnography data set, the classification of sleep stages was realize... more In this study, using the polysomnography data set, the classification of sleep stages was realized automatically with supervised learning method. In this study, sleep stage classification was carried out in three stages. In the first stage, the biomedical signal was divided into its independent components by Independent Component Analysis method. In the second stage, feature extraction was performed by using Mel Kepstrum Coefficient method. In the third stage, artificial neural networks were trained by using the extracted features and the sleep phases were estimated by using software architecture called Long ShortTerm Memory. As a result of the classification process performed in this way, the accuracy rate of the ten fold cross validation obtained for the binary classification (asleep / wake) was found to be 97,87%. For the five fold classification problem, the accuracy rate of the subject dependent algorithm and polysomnography data obtained using a single EEG channel was found to be 93,36% as a result of the ten-fold cross validation process.
2022 5th International Conference on Computing and Informatics (ICCI)
Convolutional neural networks (CNN), which have the advantage of automatically detecting the impo... more Convolutional neural networks (CNN), which have the advantage of automatically detecting the important features of the input data without any human interfere, are widely used in many applications such as face recognition, speech recognition, image classification and object detection. In real-time CNN applications, computation speed is very important as well as accuracy. However, in some applications with high computational complexity, available systems are insufficient to meet the high-speed performance demand at low power consumption. In this study, the design of the CNN accelerator hardware in FPGA is presented to meet the speed demand. In this design, CNN is considered as a streaming interface application. Thus, temporary storage amount and memory latency are reduced. Each layer is designed with maximum parallelism, taking advantage of the FPGA. Because fixed-point number representation has the advantage of low latency, it is preferred in design with negligible sacrifice of accuracy. Thus, forward propagation of a CNN can be executed at high speed in FPGA. In order to compare real-time performance, digit classification application is executed in this hardware designed in FPGA and ARM processor on the same chip. The real-time results show that the application in the hardware designed in the FPGA is 30x faster than the ARM processor.
The event related P300 potentials, positive waveforms in electroencephalography (EEG) signals, ar... more The event related P300 potentials, positive waveforms in electroencephalography (EEG) signals, are often utilized in brain computer interfaces (BCI). Many studies have been carried out to improve the performance of P300 speller systems either by developing signal processing algorithms and classifiers with different architectures or by designing new paradigms. In this study, a new paradigm is proposed for this purpose. The proposed paradigm combines two remarkable properties of being a 3D animation and utilizing column-only flashings as opposed to classical paradigms which are based on row-column flashings in 2D manner. The new paradigm is utilized in a traditional two-layer artificial neural networks model with a single output neuron, and numerous experiments are conducted to evaluate and compare the performance of the proposed paradigm with that of the classical approach. The experimental results, including statistical significance tests, are presented for single and multiple EEG e...
Erzincan Üniversitesi Fen Bilimleri Enstitüsü Dergisi, 2021
Satellite image analysis is a research area in which many research studies are carried out for ci... more Satellite image analysis is a research area in which many research studies are carried out for civil and military applications in the field of image processing. Satellite imagery has many applications including recognition, detection and classification of regions, buildings, roads, aircraft and other man-made objects. Among these, especially aircraft detection is strategically important for military applications, and forms the basis of this study. In the first phase of the study, a new dataset of aircrafts is created from Google Earth images to compensate the shortage of data set in this area. In the second stage, the detection of air vehicles was carried out using algorithms based on Convolutional Neural Network (CNN). Region-based Fully Convolutional Network (R-FCN), Single Shot Multi Box Detector (SSD) and Faster R-CNN methods are used for this process. The obtained accuracy rate for R-FCN, SSD and Faster R-CNN are 98.01%, 69.71% and 96.56%, respectively.
21st European Signal Processing Conference (EUSIPCO 2013), 2013
In this work, we have investigated the heart sound (HS) detection performance of Hidden Markov Mo... more In this work, we have investigated the heart sound (HS) detection performance of Hidden Markov Model (HMM) in respiratory sound. Respiratory sound is composed of heart sound and lung sound, and the main frequency components of these two sounds overlap with each other. To detect the locations of heart sound segments in such adverse condition accurately, the proposed method employs following steps. First, the Shannon entropy feature is extracted for robust representation of respiratory signal for different flow rates. Second, the probabilistic models are constructed by training HMM. Finally, the location of heart sound segments are efficiently estimated by the Viterbi decoding algorithm. The experimental results showed that the proposed heart sound detection method outperforms the three well-known heart sound detection methods in the literature. The average false negative rate (FNR) values for the proposed method are 5.4 ± 2.4 and 6.3 ± 1.3 for both low and medium respiratory flow rat...
2017 IEEE Global Conference on Signal and Information Processing (GlobalSIP), 2017
This paper examines filterbank parameters to extract discriminative cepstrum coefficient from EEG... more This paper examines filterbank parameters to extract discriminative cepstrum coefficient from EEG signals for sleep stage classification using well-known Support Vector Machine (SVM) algorithm. The proposed method has three main stages as feature extraction, training and classification. In feature extraction step, features are obtained using linear frequency cepstrum coefficients (LFCC) of EEG signals. Then SVM classifier is trained based on the extracted features. In the final step of classification, the class of test subject is estimated by using the trained model. Experimental results show that about an average of 95 percent correct classification rate is achievable for three classes, and this is better than the compared results available in the literature.
Abstract An efficient de-embedding air-line microwave method has been proposed for accurate relat... more Abstract An efficient de-embedding air-line microwave method has been proposed for accurate relative complex permittivity, e r = e r ′ − i e r ′ ′ , measurement of water-adulteration level within honey. It could be effectively applied to eliminate the errors arising from usage of imperfect calibration standards because it bypasses the requirement of these standards. Its accuracy is improved by utilizing the unitary and similarity properties of a passive two-port network, and then is compared with the accuracy of a calibration-dependent method present in the literature by using normalized root-mean-square-error (N-RMSE) values of e r ′ and e r ′ ′ of distilled water, in reference to the Debye value. From this comparison, it is observed that N-RMSE values calculated for e r ′ (and e r ′ ′ ) by using this calibration-dependent method and the (improved) proposed method are, respectively, around 0.1955 (0.1002) and 0.1962 (1.1067), indicating a good agreement between them. After validation the proposed de-embedding method using distilled water measurements, tested pure honey was adulterated with distilled water by different percentage values δ ranging from 1% to 10% in 1% increments. It is observed that the maximum distance between extracted e r ′ (or e r ′ ′ ) values of adulterated honey by the applied calibration-dependent method and the proposed method is less than 2%. Afterwards, an empirical formula was devised to fit e r ′ and e r ′ ′ values from measured e r of water-adulterated honey with various δ levels. It is noted that extracted e r ′ is much more better fitted than extracted e r ′ ′ , especially for δ ≤ 4 . Next, an optimization process is followed to evaluate the frequency for optimum prediction of adulteration levels using the empirical formula based on e r ′ or e r ′ ′ . It is noticed that optimized δ values using the empirical formula based on e r ′ (with an average prediction error of around 0.071 at 4.5 GHz) are superior to optimized δ values using the empirical formula based on e r ′ ′ (with an average prediction error of around 0.085 at 4.2 GHz) for prediction of previously known δ values. Sensitivity and uncertainty analyses were performed to assess and improve the accuracy of the proposed method.
Porous silicon (PSi) Fabry-Pérot (FP) cavities are the sensor types widely employed in sensing ap... more Porous silicon (PSi) Fabry-Pérot (FP) cavities are the sensor types widely employed in sensing applications of chemical, biological, or gas molecules. Prior to sensor operation, each fabricated empty (no-molecule) PSi FP cavity is characterized by their optical properties (refractive indices and thicknesses of each layer). For this characterization, a scanning electron microscope (SEM) image of a cut-sample from the batch FP cavity is measured to monitor the thickness variation of each layer. This technique is surely destructive and gives local information of only the broken cut-sample. In this Letter, we apply the full spectra fitting technique for nondestructive and accurate optical characterization of empty PSi FP cavities. To demonstrate the potential of this technique, we obtained the optimized thicknesses of each layer of two fabricated PSi FP cavities with resonating wavelengths of 542 nm and 1456 nm and compared them with the thicknesses obtained from their SEM images. From this analysis, we note that the proposed technique can be a good candidate for nondestructive characterization of empty FP cavities.
2015 23nd Signal Processing and Communications Applications Conference (SIU), 2015
In this study, it is aimed to produce microelectrodes which can be used in the detection of neuro... more In this study, it is aimed to produce microelectrodes which can be used in the detection of neurotransmitters that are related with brain disorders such as Parkinson, Epilepsy, and Schizophrenia and that exist in the central nervous system (CNS). A 4-channel, ceramic-based fabrication is performed towards this goal by using photolithographic methods. The time-current graphic response against the addition of H2O2 the produced microelectrode is analyzed in the calibration test. It is observed that the response is in stepwise form. In addition, limit of detection (LOD) of the produced microelectrodes and linearity values are shown to be within the desired ranges.
Medical & Biological Engineering & Computing, 2014
This work presents a detailed framework to detect the location of heart sound within the respirat... more This work presents a detailed framework to detect the location of heart sound within the respiratory sound based on temporal fuzzy c-means (TFCM) algorithm. In the proposed method, respiratory sound is first divided into frames and for each frame, the logarithmic energy features are calculated. Then, these features are used to classify the respiratory sound as heart sound (HS containing lung sound) and non-HS (only lung sound) by the TFCM algorithm. The TFCM is the modified version fuzzy c-means (FCM) algorithm. While the FCM algorithm uses only the local information about the current frame, the TFCM algorithm uses the temporal information from both the current and the neighboring frames in decision making. To measure the detection performance of the proposed method, several experiments have been conducted on a database of 24 healthy subjects. The experimental results show that the average false-negative rate values are 0.8 ± 1.1 and 1.5 ± 1.4 %, and the normalized area under detection error curves are 0.0145 and 0.0269 for the TFCM method in the low and medium respiratory flow rates, respectively. These average values are significantly lower than those obtained by FCM algorithm and by the other compared methods in the literature, which demonstrates the efficiency of the proposed TFCM algorithm. On the other hand, the average elapsed time of the TFCM for a data with length of 0.2 ± 0.05 s is 0.2 ± 0.05 s, which is slightly higher than that of the FCM and lower than those of the other compared methods.
This paper investigates the utility of a likelihood ratio test (LRT) combined with an efficient a... more This paper investigates the utility of a likelihood ratio test (LRT) combined with an efficient adaptation procedure for the purpose of detecting the heart sound (HS) with lung sound and the lung sound only (non-HS) segments in a respiratory signal. The proposed detection method has four main stages: feature extraction, training of the models, detection, and adaptation of the model parameter. In the first stage, the logarithmic energy features are extracted for each frame of respiratory sound. In the second stage, the probabilistic models for HS and non-HS segments are constructed by training Gaussian mixture models (GMMs) with an expectation maximization algorithm in a subject-independent manner, and then the HS and non-HS segments are detected by the results of the LRT based on the GMMs. In the adaptation stage, the subject-independent trained model parameter is modified online using the observed test data to fit the model parameter of the target subject. Experiments were performe...
We propose a novel method for identification of various gases using combined measurements of wave... more We propose a novel method for identification of various gases using combined measurements of wavelength shift (λ 0) and reflectivity difference (R). The method relies on plotting the measurement data on the R-λ 0 space and exploits in the identification process the effect of significant loss behavior of bulk silicon near or within the visible range (for lossless gas vapors). It has been validated by reflectivity measurements on a Fabry-Pérot cavity resonating around 745 nm for four different gas vapors whose reflective indices are close to each other and by two metrics for separability/identification analysis by the scattering matrix method. It is noted that the combined measurements of λ 0 and R result in better identification than that obtained by the measurement of λ 0 (or R) itself.
Three different techniques are applied for accurate constitutive parameters determination of isot... more Three different techniques are applied for accurate constitutive parameters determination of isotropic split-ring resonator (SRR) and SRR with a cut wire (Composite) metamaterial (MM) slabs. The first two techniques use explicit analytical calibrationdependent and calibration-invariant expressions while the third technique is based on Lorentz and Drude dispersion models. We have tested these techniques from simulated scattering (S-) parameters of two classic SRR and Composite MM slabs with various level of losses and different calibration plane factors. From the comparison, we conclude that whereas the extracted complex permittivity of both slabs by the analytical techniques produces unphysical results at resonance regions, that by the dispersion model eliminates this shortcoming and retrieves physically accurate constitutive parameters over the whole analyzed frequency region. We argue that incorrect retrieval of complex permittivity by analytical methods comes from spatial dispersion effects due to the discreteness of conducting elements within MM slabs which largely vary simulated S-parameters in the resonance regions where the slabs are highly spatially dispersive.
Heart sounds are the main obstacle in lung sound analysis. To tackle this obstacle, we propose a ... more Heart sounds are the main obstacle in lung sound analysis. To tackle this obstacle, we propose a diagnosis algorithm that uses singular spectrum analysis (SSA) and frequency features of heart and lung sounds. In particular, we introduce a frequency coefficient that shows the frequency difference between heart and lung sounds. The proposed algorithm is applied to a synthetic mixture of heart and lung sounds. The results show that heart sounds can be extracted successfully and localizations for the first and second heart sounds are remarkably performed. An error analysis of the localization results shows that the proposed algorithm has fewer errors compared to the SSA method, which is one of the most powerful methods in the localization of heart sounds. The presented algorithm is also applied in the cases of recorded respiratory sounds from the chest walls of five healthy subjects. The efficiency of the algorithm in extracting heart sounds from the recorded breathing sounds is verified with power spectral density evaluations and listening. Most studies have used only normal respiratory sounds, whereas we additionally use abnormal breathing sounds to validate the strength of our achievements.
Engineering Science and Technology, an International Journal, 2022
Drone detection and classification, important in military and civilian applications, are performe... more Drone detection and classification, important in military and civilian applications, are performed using different sensor signals. Proposed study handles this task using Radio Frequency (RF) signals utilizing basic machine learning methods. It is composed of two main stages as feature extraction succeeded by training/testing of the model. In feature extraction stage, valuable information for classification, contained in the RF signal, is obtained. For this purpose, spectral features, frequently used in speech processing applications, are employed. Specifically, Power Spectral Density (PSD), Mel-Frequency Cepstral Coefficients (MFCC) and Linear Frequency Cepstral Coefficients (LFCC) are adopted by adjusting filter bank margins and parameters for this task. In the second stage, a Support Vector Machine (SVM) classifier is first trained based on the obtained features and finally tested to measure its performance. All experimental studies are carried out using publicly available DroneRF dataset. This dataset contains 2-Class, 4-Class and 10-Class samples for drone existence vs. background (BG), drone types and drone operation modes, respectively. The best classification results are obtained using, PSD, MFCC and LFCC based features for 2-Class, MFCC and LFCC based features for 4-Class and LFCC based features for 10-Class. Accuracy rates for 2-Class, 4-Class and 10-Class are 100%, 98.67% and 95.15%, respectively. These results show that the proposed method outperforms the results given in the literature for DroneRF dataset.
Engineering Science and Technology, an International Journal, 2022
Drone detection and classification, important in military and civilian applications, are performe... more Drone detection and classification, important in military and civilian applications, are performed using different sensor signals. Proposed study handles this task using Radio Frequency (RF) signals utilizing basic machine learning methods. It is composed of two main stages as feature extraction succeeded by training/testing of the model. In feature extraction stage, valuable information for classification, contained in the RF signal, is obtained. For this purpose, spectral features, frequently used in speech processing applications, are employed. Specifically, Power Spectral Density (PSD), Mel-Frequency Cepstral Coefficients (MFCC) and Linear Frequency Cepstral Coefficients (LFCC) are adopted by adjusting filter bank margins and parameters for this task. In the second stage, a Support Vector Machine (SVM) classifier is first trained based on the obtained features and finally tested to measure its performance. All experimental studies are carried out using publicly available DroneRF dataset. This dataset contains 2-Class, 4-Class and 10-Class samples for drone existence vs. background (BG), drone types and drone operation modes, respectively. The best classification results are obtained using, PSD, MFCC and LFCC based features for 2-Class, MFCC and LFCC based features for 4-Class and LFCC based features for 10-Class. Accuracy rates for 2-Class, 4-Class and 10-Class are 100%, 98.67% and 95.15%, respectively. These results show that the proposed method outperforms the results given in the literature for DroneRF dataset.
Uploads
Papers by yucel ozbek