Yazar "Gürkan, Hakan" seçeneğine göre listele
Listeleniyor 1 - 15 / 15
Sayfa Başına Sonuç
Sıralama seçenekleri
Öğe A novel biometric identification system based on fingertip electrocardiogram and speech signals(Academic Press, 2021) Guven, Gokhan; Guz, Umit; Gürkan, HakanIn this research work, we propose a one dimensional Convolutional Neural Network (CNN) based biometric identification system that combines speech and ECG modalities. The aim is to find an effective identification strategy while enhancing both the confidence and the performance of the system. In our first approach, we have developed a voting-based ECG and speech fusion system to improve the overall performance compared to the conventional methods. In the second approach, we have developed a robust rejection algorithm to prevent unauthorized access to the fusion system. We also presented a newly developed ECG spike and inconsistent beats removal algorithm to detect and eliminate the problems caused by portable fingertip ECG devices and patient movements. Furthermore, we have achieved a system that can work with only one authorized user by adding a Universal Background Model to our algorithm. In the first approach, the proposed fusion system achieved a 100% accuracy rate for 90 people by taking the average of 3-fold cross-validation. In the second approach, by using 90 people as genuine classes and 26 people as imposter classes, the proposed system achieved 92% accuracy in identiying genuine classes and 96% accuracy in rejecting imposter classes.Öğe A YOLOv3-Based Smart City Application For Children’s Playgrounds(Bursa Teknik Üniversitesi, 2021) İnkaya, Mehmet Fatih; Gürkan, HakanAccording to the reports of Public Health Institution, approximately 250,000 rabies-risk animal bites occur per year in Turkey. Most of these bites are caused by dogs and most of the victims are the children who play in playgrounds. With the development of deep learning-based computer vision technology, autonomous detection of dangerous objects (handguns, knives, dogs, etc.) in these children’s playgrounds has become a crucial security application. In this paper, a real-time dog detection model based on YOLOv3 deep learning algorithm is proposed as a new smart city security application and this model is applied to the selected children’s playground. Firstly, in view of the problem of insufficient stray dog image data in the original datasets, new images of stray dogs have been taken from an animal shelter and they have been added to the dataset. These new images have effectively enriched the diversity of training data and improved the training performance of the proposed model. The proposed model has been optimized by utilizing various hyperparameters and the results have been compared with each other. The model with the best evaluation scores is proposed and applied to detect dogs automatically by the fast emergency station located in the selected children’s playground. The real-time application has achieved 82.59% of AP with adjusted hyperparameters.Öğe Biometric identification using fingertip electrocardiogram signals(Springer London Ltd, 2018) Guven, Gokhan; Gürkan, Hakan; Guz, UmitIn this research work, we present a newly fingertip electrocardiogram (ECG) data acquisition device capable of recording the lead-1 ECG signal through the right- and left-hand thumb fingers. The proposed device is high-sensitive, dry-contact, portable, user-friendly, inexpensive, and does not require using conventional components which are cumbersome and irritating such as wet adhesive Ag/AgCl electrodes. One of the other advantages of this device is to make it possible to record and use the lead-1 ECG signal easily in any condition and anywhere incorporating with any platform to use for advanced applications such as biometric recognition and clinical diagnostics. Furthermore, we proposed a biometric identification method based on combining autocorrelation and discrete cosine transform-based features, cepstral features, and QRS beat information. The proposed method was evaluated on three fingertip ECG signal databases recorded by utilizing the proposed device. The experimental results demonstrate that the proposed biometric identification method achieves person recognition rate values of 100% (30 out of 30), 100% (45 out of 45), and 98.33% (59 out of 60) for 30, 45, and 60 subjects, respectively.Öğe Compression of ECG Signals Using Long Short-Term Memory based Sequence-to-Sequence Autoencoder(Institute of Electrical and Electronics Engineers Inc., 2020) Aydemir, Gürkan; Bekiryazıcı, Tahir; Gürkan, HakanThis study proposes a novel long short-term memory based sequence-to-sequence autoencoder model to compress ECG signals. The efficiency of this new method is illustrated on MIT-BIH Arrhythmia dataset. In the conducted experiments, the proposed architecture achieves %21.14 mean-independent percentage mean square difference (MPRD) with a constant compression ratio value of 33 : 1. © 2020 IEEE.Öğe Compression of the biomedical images using quadtree-based partitioned universally classified energy and pattern blocks(Springer London Ltd, 2019) Gezer, Murat; Gargari, Sepideh Nahavandi; Guz, Umit; Gürkan, HakanIn this work, an efficient low bit rate image coding/compression method based on the quadtree-based partitioned universally classified energy and pattern building blocks (QB-UCEPB) is introduced. The proposed method combines low bit rate robustness and variable-sized quantization benefits of the well-known classified energy and pattern blocks (CEPB) method and quadtree-based (QB) partitioning technique, respectively. In the new method, first, the QB-UCEPB is constructed in the form of variable length block size thanks to the quadtree-based partitioning rather than fixed block size partitioning which was employed in the conventional CEPB method. The QB-UCEPB is then placed to the transmitter side as well as receiver side of the communication channel as a universal codebook manner. Every quadtree-based partitioned block of the input image is encoded using three quantities: image block scaling coefficient, the index number of the QB-UCEB and the index number of the QB-UCPB. These quantities are sent from the transmitter part to the receiver part through the communication channel. Then, the quadtree-based partitioned input image blocks are reconstructed in the receiver part using a decoding algorithm, which exploits the mathematical model that is proposed. Experimental results show that using the new method, the computational complexity of the classical CEPB is substantially reduced. Furthermore, higher compression ratios, PSNR and SSIM levels are achieved even at low bit rates compared to the classical CEPB and conventional methods such as SPIHT, EZW and JPEG2000.Öğe ECG Biometric Identification Method based on Parallel 2-D Convolutional Neural Networks(Bursa Teknik Üniversitesi, 2019) Hanılcı, Ayca; Gürkan, HakanIn this paper, an ECG biometric identification method, based on a two-dimensional convolutional neural network, is introduced for biometric applications. The proposed model includes two-dimensional convolutional neural networks that work parallel and receive two different sets of 2-dimensional features as input. First, ACDCT features and cepstral properties are extracted from overlapping ECG signals. Then, these features are transformed from one-dimensional representation to two-dimensional representation by matrix manipulations. For feature learning purposes, these two-dimensional features are given to the inputs of the proposed model, separately. Finally, score level fusion is applied to identify the user. Our experimental results show that the proposed biometric identification method achieves an accuracy of %88.57 and an identification rate of 90.48% for 42 persons.Öğe ECG Compression method based on convolutional autoencoder and discrete wavelet transform(Institute of Electrical and Electronics Engineers Inc., 2020) Bekiryazıcı, Tahir; Gürkan, HakanIn this work, a compression method based on one dimensional convolutional autocoder architecture and wavelet transform is proposed for the compression of ECG signals. The proposed method is tested on the MIT-BIH Arrhythmia Database and its performance is evaluated with respect to compression ratio (CR) and mean-independent percentage mean square difference (MPRD). Experimental results showed that the proposed method achieves an average CR value of 32.27:1 with an averages MPRD of %18.91. © 2020 IEEE.Öğe ECG Signal Compression Based on Wavelet Packet Transform and Residual Vector Quantization(Institute of Electrical and Electronics Engineers Inc., 2025) Beki?Ryazici, Tahir; Aydemir, Gürkan; Gürkan, HakanEfficient compression of electrocardiogram (ECG) signals is of great importance, particularly in facilitating data transmission and reducing storage costs in remote monitoring systems. For this purpose, a two-stage compression method based on Wavelet Packet Transform (WPT) and Residual Vector Quantization (RVQ) has been developed. The performance of the proposed method was tested on the widely used MIT-BIH Arrhythmia dataset. Performance evaluation was conducted using statistical metrics such as compression ratio (CR), percentage root-mean-square difference normalized (PRDN), and quality score (QS). Experimental results show that the method achieved a compression ratio of 8.8 : 1 with 6.54% PRDN and 17.6 : 1 with 14.38% PRDN. © 2025 IEEE.Öğe Effective semi-supervised learning strategies for automatic sentence segmentation(Elsevier Science Bv, 2018) Dalva, Dogan; Guz, Umit; Gürkan, HakanThe primary objective of sentence segmentation process is to determine the sentence boundaries of a stream of words output by the automatic speech recognizers. Statistical methods developed for sentence segmentation requires a significant amount of labeled data which is time-consuming, labor intensive and expensive. In this work, we propose new multi-view semi-supervised learning strategies for sentence boundary classification problem using lexical, prosodic, and morphological information. The aim is to find effective semi-supervised machine learning strategies when only small sets of sentence boundary labeled data are available. We primarily investigate two semi-supervised learning approaches, called self-training and co-training. Different example selection strategies were also used for co-training, namely, agreement, disagreement and self-combined. Furthermore, we propose three-view and committee-based algorithms incorporating with agreement, disagreement and self-combined strategies using three disjoint feature sets. We present comparative results of different learning strategies on the sentence segmentation task. The experimental results show that the sentence segmentation performance can be highly improved using multi-view learning strategies that we proposed since data sets can be represented by three redundantly sufficient and disjoint feature sets. We show that the proposed strategies substantially improve the average baseline F-measure of 67.66% to 75.15% and 64.84% to 66.32% when only a small set of manually labeled data is available for Turkish and English spoken languages, respectively. (c) 2017 Elsevier B.V. All rights reserved.Öğe Evrişimsel sinir ağı ve QRS imgeleri kullanarak EKG tabanlı biyometrik tanıma yöntemi(2020) Gürkan, Hakan; Hanilçi, AyçaMedikal uygulamalarda yaygın olarak kullanılan elektrokardiyogram (EKG) işaretleri, aldatma saldırılarına karşı güçlü kılan yaşam işareti olma özelliği sayesinde, biyometrik uygulamalar için bir biyometrik büyüklük olarak kullanılmaya başlanmıştır. Bilgisayar sistemlerinin hesaplama güçlerinin artmasına bağlı olarak kişi tanıma ve sınıflandırma doğruluğunu arttırmak amacıyla son yıllarda EKG biyometrik tanıma için birkaç evrişimsel sinir ağı (ESA) tabanlı yöntem sunulmuştur. Bu çalışmada, QRS (QRS dalgası) imgeleri ve 2 boyutlu ESA yapısı kullanılarak EKG işaretleri tabanlı bir biyometrik tanıma yöntemi önerilmiştir. Önerilen yöntemde, ilk olarak EKG işaretleri gürültü temizleme ve QRS belirleme algoritmalarından geçirilerek QRS bölütlerine ayrılmıştır. Elde edilen bu bölütler R noktalarına göre hizalandıktan sonra 256x256 büyüklüğünde QRS imgesi olarak adlandırılan 2 boyutlu EKG işaretlerine dönüştürülmüştür. Son olarak elde edilen bu QRS imgelerinin giriş olarak uygulandığı 2 boyutlu bir ESA modeli geliştirilerek biyometrik tanıma gerçekleştirilmiştir. Önerilen yöntemin başarımı diğer ESA tabanlı EKG biyometrik tanıma yöntemleri ile karşılaştırmalı olarak incelenmiştir. Önerilen yöntem 46 kişiden oluşan bir EKG veri kümesi üzerinde %98.08 doğruluk oranı ve %99.275 kişi tanıma oranı sağlamıştır.Öğe Extension of Conventional Co-Training Learning Strategies to Three-View and Committee-Based Learning Strategies for Effective Automatic Sentence Segmentation(Institute of Electrical and Electronics Engineers Inc., 2018) Dalva, Dogan; Güz, Ümit; Gürkan, HakanThe objective of this work is to develop effective multiview semi-supervised machine learning strategies for sentence boundary classification problem when only small sets of sentence boundary labeled data are available. We propose three-view and committee-based learning strategies incorporating with co-training algorithms with agreement, disagreement, and self-combined learning strategies using prosodic, lexical and morphological information. We compare experimental results of proposed three-view and committee-based learning strategies to other semi-supervised learning strategies in the literature namely, self-training and co-training with agreement, disagreement, and self-combined strategies. The experiment results show that sentence segmentation performance can be highly improved using multi-view learning strategies that we propose since data sets can be represented by three redundantly sufficient and disjoint feature sets. We show that the proposed strategies substantially improve the average performance when only a small set of manually labeled data is available for Turkish and English spoken languages, respectively. © 2018 IEEE.Öğe EXTENSION OF CONVENTIONAL CO-TRAINING LEARNING STRATEGIES TO THREE-VIEW AND COMMITTEE-BASED LEARNING STRATEGIES FOR EFFECTIVE AUTOMATIC SENTENCE SEGMENTATION(Ieee, 2018) Dalva, Dogan; Guz, Umit; Gürkan, HakanThe objective of this work is to develop effective multi-view semi-supervised machine learning strategies for sentence boundary classification problem when only small sets of sentence boundary labeled data are available. We propose three-view and committee-based learning strategies incorporating with co-training algorithms with agreement, disagreement, and self-combined learning strategies using prosodic, lexical and morphological information. We compare experimental results of proposed three-view and committee-based learning strategies to other semi-supervised learning strategies in the literature namely, self-training and co-training with agreement, disagreement, and self-combined strategies. The experiment results show that sentence segmentation performance can be highly improved using multi-view learning strategies that we propose since data sets can be represented by three redundantly sufficient and disjoint feature sets. We show that the proposed strategies substantially improve the average performance when only a small set of manually labeled data is available for Turkish and English spoken languages, respectively.Öğe Lightweight CNN-Based Intrusion Detection for Automotive CAN Bus in Light Commercial Vehicles(Bursa Teknik Üniversitesi, 2025) Tüfekcioğlu, Emre; Hanilçi, Cemal; Gürkan, HakanWith the rapid advancement of digitalization and automation, modern vehicles, especially in the light commercial segment, have evolved into complex, interconnected platforms resembling mobile computing systems. This transformation has increased the dependency on in-vehicle communication networks and, as a result, exposed them to a wider range of cybersecurity threats. A fundamental aspect of the proposed method is the use of a lightweight CNN model specific for deployment in embedded automotive environments with limited computational resources and optimized for efficiency. Operating on low-power hardware platforms such as edge ECUs, the tiny device developed in this study works effectively unlike conventional deep learning architectures seeking high processing power and memory. Despite its minimal computational footprint, the model is capable of accurately distinguishing between legitimate and spoofed communication traffic, as well as detecting a variety of attack forms that target different CAN protocol components. The performance metrics of the model further highlight its effectiveness, achieving a ROC AUC Score of 0.9887, an Accuracy of 0.9887, a Precision of 0.9825, a Recall of 0.9952, and an F1-Score of 0.9888. Particularly for real-time on-vehicle intrusion detection systems, this harmony between performance and efficiency makes the strategy especially important. Just as importantly is the introduction of a specifically produced hybrid dataset, which is fundamental for system evaluation and training. The dataset aggregates synthetic generated attack scenarios with real-world spoofing, injection, and denial-of- service (DoS) conditions using actual CAN traffic acquired from a J1939-compliant light commercial vehicle. Standard 11-bit identities combined with industrial communication protocols help the dataset to reflect real-world vehicle dynamics across several ECUs under various scenarios. The model can learn fine-grained patterns often missed by conventional rule-based or manually engineered approaches by means of the image-like transformation of CAN messages—preserving bit-level and temporal information. In intelligent transportation systems, the lightweight CNN architecture and the strong dataset combine to create a scalable and deployable IDS framework that can improve in-vehicle cybersecurity.Öğe Measurement of Geometric Dimensions of Hexagonal Type Washer Parts Using Machine Vision-Based Systems(Institute of Electrical and Electronics Engineers Inc., 2024) Poyraz, Ahmet Gökhan; Dirik, Ahmet Emir; Gürkan, HakanIn this study, a method is proposed for measuring the geometric properties of hexagonal precision adjustment washers. The method primarily involves measuring the hypothetical diameter that encompasses the outermost part of the hexagon and the key slot dimensions of the hexagon. For the key slot dimension, subpixel-based edge detection is performed following a preprocessing step. Subsequently, the enclosing circle and diameter are estimated based on the detected points. For the key slot measurement, the first step involves Hough transform-based line detection. By predicting the equations of the detected lines, the distances between the opposing lines are computed. According to the results, both the proposed diameter measurement algorithm and the key slot measurement algorithm function with an average error of approximately 13 ?m in the utilized dataset. When considering part tolerances and average error amounts, it is concluded that the proposed algorithms are successful and feasible for practical use. © 2024 IEEE.Öğe Sub-Pixel counting based diameter measurement algorithm for industrial Machine vision(Elsevier B.V., 2024) Poyraz, Ahmet Gökhan; Kaçmaz, Mehmet Ali; Gürkan, Hakan; Dirik, Ahmet EmirIn recent years, there has been a notable surge in the utilization of industrial image processing applications across various sectors, including automotive, medical, and space industries. These applications rely on specialized camera systems and advanced image processing techniques to accurately measure working products with precise tolerances. This research presents a novel fast algorithm for measuring the diameter of a ring, employing a subpixel counting method. The algorithm classifies image pixels into two categories: full pixels and transition pixels. Full pixels reside entirely within the inner region of the workpiece, while transition pixels represent gray pixels that reside at the boundary between the workpiece and its background. To ensure accurate determination of the object area, the proposed method incorporates normalization to account for the contribution of transition pixels alongside full pixels. Subsequently, the circle area equation is employed to calculate the diameter. Moreover, a robust threshold selection method is introduced to effectively distinguish pixels with gray intensities. The experimental setup consists of an industrial camera equipped with telecentric lenses and appropriate illumination. The results demonstrate that the proposed algorithm achieves a 3–10 % improvement in accuracy compared to existing approaches. In terms of measuring sensitivity, the operational sensitivity of the proposed methodology is quantified as 1/20th of the pixel size, exhibiting an average uncertainty of 1 µm. Furthermore, the proposed method surpasses existing works by at least 12.5 % to 35 % in terms of benchmarking computing time. © 2023 Elsevier Ltd
