Search

28th Signal Processing and Communications Applications Conference, SIU 2020 -- 5 October 2020 through 7 October 2020 -- -- 166413 2-s2.0-85100298433 This paper introduces a design of a single rectangular Microstrip Patch Antenna (MPA) with inset feed at frequency of 60 GHz using a Frequency Selective Surface (FSS). After implanting the FSS in the single MPA, it is found that some parameters have been improved. The selected geometry was the Square Loop (SL) element in FSS which has the best performance over other simple shapes. This paper aims to introduce an improvement of the parameter of a MPA, particularly, the gain, return loss, and bandwidth. These improvements has been investigated using Computer Simulation Technology (CST) Microwave Studio (MWS) simulator. The design is performed for 60 GHz applications. The 60 GHz band suffers from severe attenuation due to propagation. The oxygen absorption and the free-space path loss happen at 60 GHz are presumed in this study. These combined factors result in a poor link budget. © 2020 IEEE.

4th International Conference on Computer Science and Engineering, UBMK 2019 -- 11 September 2019 through 15 September 2019 -- -- 154916 It is most common way to persue the attendance of student is signed paper that every student sign it with their names. It causes some problems such as losing the paper in time and fraud that some student sign it in the name of their absent friends. That project aims to prevent these kind of problems with using the method of fingerprint. It enables that taking student information with the using of fingerprint sensor to cary web server and take these information from a web site by ZigBee protocol. The controller designed as a reader of fingerprint in the system. The web site designed to register of students and to monitoring all of them. All information about the web sites and the data base are saved in the web server. The flow of information between the controller and web server are provided by ZigBee protocol according to system aim. All information of students fingerprints are saved In The System of Student Attendance. Teachers are able to start attendance of existing students with the filling of requirement information after enter the system and also take reports about past attendace lists in Whenever in every place. Teachers are able to start attendancy of existing students with the filling of requirement information after enter the system and also take reports about past attendace lists in Whenever in every place. If the match is successful, student is accepted to participate the class and automaticly saved in data base. © 2019 IEEE.

27th Signal Processing and Communications Applications Conference, SIU 2019 -- 24 April 2019 through 26 April 2019 -- -- 151073, In this paper, end to end average bit error rate (BER) for three-hop cooperative communication systems with a decode-and-forward (DF) relay is derived in the closed-form in the presence of error propagation. The derived end-to-end BER expression is verified via computer simulations. It is shown that the threshold selection for the relays has dominant effect on the error performance of the system. In this paper, we multi-hop communication between base station and cell-edge users to achieve ultra-wide coverage which is one of the important requirements for 5G and beyond. © 2019 IEEE.

EgeUn###, Internet of Things (IoT) can be defined as the communication network that connects physical devices such as sensors, actuators, electronic home appliances and smart devices with each other and other systems. As part of IoT, numerous physical devices work concurrently and produce data continuously with big volumes. In this regard, NoSQL technology is utilized, because relational databases are insufficient to store and query the IoT data. NoSQL is the database technology that allows rapid organization of different non-relational data types. NoSQL databases are divided into four categories, which are key-value based, column-oriented, document-oriented and graph databases. Since graph databases do not provide effective solutions for IoT data, several studies in the literature focus on the use of database technologies in the other three categories for the purpose of storing and querying IoT data more effectively. In this paper, it is aimed to evaluate the storage and query performance of MongoDB on IoT data. It has been decided to use MongoDB, which is a document-oriented NoSQL database technology, because of its popular usage and high performance. Yahoo Cloud Serving Benchmark (YCSB), which is an open source tool developed by Yahoo that allows database operations at different volumes, is used for performance measurements. YCSB has been made compatible with the IoT dataset that is used for performance measurements, by re-organizing its DB Interface Layer for this study. The paper is concluded with discussion of the performance measurement results. © 2018 IEEE.

2nd International Conference on Computer Science and Engineering, UBMK 2017 --5 October 2017 through 8 October 2017 -- -- In optimization problems, Genetic Algorithms are one of the most commonly used methods to search optimum points of a given function. These algorithms stochastically select the individual that is close to the optimum point in the population. By choosing appropriate individual in each iteration, it is desired to find the best individual step by step, or converge to the best individual. Therefore, it is significantly important to have a decent selection method in genetic algorithms. In this paper, it is aimed to improve Aggressive and Integrated Aggressive Selection methods which were already proposed. The performance of the improved methods that are proposed in this paper are compared with aggressive selection methods and integrated aggressive selection methods, as well as, most commonly used standard selection methods; Roulette Wheel, Linear Ranking and Tournament. It is observed that results of improved methods are predominant comparing to the other algorithms. © 2017 IEEE.

Spatial modulation (SM) is a simple multi-input multi-output (MIMO) communication technique in which a single transmit antenna is active for each symbol period. Quadrature spatial modulation (QSM), on the other hand, is a form of SM in which independent transmit antenna selection is made for the real and imaginary parts of the symbols. In these techniques, the spectral efficiency (SE) can be increased logarithmically by increasing the number of antennas used on the transmitter-side. Besides, in classical SM and QSM schemes, the number of antennas must be the power of two. In this paper, a flexible QSM scheme is proposed, which eliminates the necessity of antenna number and increases SE value linearly with the number of antennas for RF systems. This scheme, called hardware-aware quadrature spatial modulation (HAQSM), uses optimized constellations for the desired SE value. Optimal constellations are obtained using convex optimization under a constraint aimed at keeping the distance between neighboring symbols above the desired value. The bit error rate (BER) performances of the HAQSM scheme were obtained by Monte Carlo simulations and compared with the existing schemes in the literature.

6th International Conference on Computer Science and Engineering, UBMK 2021 -- 15 September 2021 through 17 September 2021 -- -- 176826, Nowadays, shopping is done more comfortably and without time constraints with the throwing of e-commerce platforms. These platforms allow consumers to examine reviews before purchasing products. Thus, consumers can decide whether to buy a product with positive or negative comments about the products. In this paper, Turkish sentiment analysis was carried out on the product comments at the Hepsiburada platform. For sentiment analysis, firstly, the success of Random Forest, Naive Bayes and Logistic Regression machine learning methods was measured. Then, the effect of BERT, ELECTRA and ALBERT language models on sentiment analysis was analyzed and the success of language models was compared with machine learning methods. While Naive Bayes achieved the highest accuracy with 89.95% among machine learning methods, ELECTRA was the most successful with 92.54% among language models. As a result of the study, it has been shown that the ELECTRA and ALBERT language models are more successful than machine learning methods. © 2021 IEEE

28th Signal Processing and Communications Applications Conference, SIU (2020; Gaziantep, Turkey), In this paper, we propose an automatic text summarization model for Turkish news articles using machine learning models. Our proposed model uses sentence position, speech expression, presence of named entities and statements, term frequency and title similarity as features. We construct and share a new dataset for Turkish text summarization. In our experiments, we show that all our features we use have a positive impact on the performance of the system. In addition, we show that our model outperforms the latent semantic analysis based baseline method.

14th Turkish National Software Engineering Symposium, UYMS 2020 -- 7 October 2020 through 9 October 2020 -- -- 164914, This paper evaluates the use of Spine animation library during industrial mobile game software development. The evaluation results showed that the use of both Spine and its supporting tools provides the rapid application of the same animation to various characters, shortens the transfer of animation constructs between animators and game developers and increases the production of required software approximately seven times. However, we also observed that the library was incapable of properly reflecting some of the light events inside games and caused problems in viewing the two animations together. © 2020 IEEE.

14th Turkish National Software Engineering Symposium, UYMS 2020; Istanbul; Turkey; 7 October 2020 through 9 October 2020, Günümüzde esnek, güvenilir ve duyarlı (responsive) yazılımlara olan talebin artması ve bu talebe karşılık verebilen internet altyapısının olması yeni nesil bulut uygulamalarının gelişmesine öncülük etmiştir. Bunun sonucunda, bulut tabanlı dağıtık web uygulamalarının birlikte çalışarak oluşturduğu mikroservis tabanlı mimari popülerlik kazanmıştır. Mikroservis tabanlı mimari oldukça yeni olup bu mimariyi kullanarak yazılım geliştiren firmaların kültürlerini değiştirme gerekliliği doğmuştur. Ancak, literatürde mikroservis tabanlı mimaride analiz ve tasarım konusunda uygulanan yöntemleri ele alan çalışma sayısı çok azdır. Bu çalışmada, Türkiye’deki yazılım organizasyonlarının mikroservis tabanlı proje geliştirirken başvurdukları analiz ve tasarım yöntemlerini ele alan bir anket düzenlenmiştir. Anket sonuçları, mikroservis tabanlı proje konusunda tecrübesi olan yazılım organizasyonlarının analiz ve tasarım konusundaki bakış açılarını ortaya çıkarmaktadır. Elde edilen sonuçlar mikroservis kullanan organizasyonlar ile ilgili genel tabloyu göstermekte ve araştırmacılar için çalışma yapılabilecek konuları önermektedir., The demand for more flexible, responsive and reliable software applications and the availability of internet infrastructure to respond to this demand, led to a new generation of cloud-based web applications. As a result, cloud- based distributed web applications working together in a microservice-based architecture has gained popularity. The concept of microservice-based architecture is quite new and software organizations need to transform their culture to develop applications in this fashion. However, there is a lack of research studies in the literature that explores the common practices for the analysis and design of microservices. Thus, we performed a survey to explore the organizational choices on software analysis and design when working with microservices. In this paper, we present the results from Turkey. The results provide a snapshot of the software industry that utilizes microservices and suggests a set of challenges researchers can focus on in the area.

Bu bildiri, derin öğrenme yöntemleri uygulayarak uzaktan algılamalı optik görüntülerde bina yoğunluğunun noktasal olarak kestirilmesi ile ilgilidir. Bu çalışma kapsamında, evrişimsel sinir ağına (ESA) dayalı derin öğrenme yöntemlerine başvurulmuştur. Önceden eğitilmiş, VGG-16 ve FCN-8s derin mimarileri bu probleme uyarlanmış ve ince ayar verilerek eğitilmiştir. Kestirilen değerler yerleşim bölgelerinde bina yoğunluk haritası oluşturmak için kullanılmıştır. Her iki mimarinin karşılaştırmalı benzetim sonuçları, güdümlü eğitim için binaları gösteren detaylı haritalara ihtiyaç duyulmadan hassas yoğunluk kestirimi yapılabileceğini göstermektedir. This paper is about point-wise estimation of building density from remote sensing optical imagery using deep learning methods. Convolutional neural network (CNN) based deep learning approaches are used for this work. Pre-trained VGG-16 and FCN-8s deep architectures are adapted to the problem and fine-tuned with additional training. Estimated values are used to generate building heat maps in urban areas. Comparative simulation results of the two architectures reveal that accurate density estimation is possible without the need for detailed maps of building locations during supervised training. IEEE IEEE Turkey Sect Publisher's Version WOS:000609879900080

Kelime ağı, anlamlarına göre organize edilmiş kelimeleri barındıran bir veritabanıdır. Bir kelime ağı, sahip olduğu kelimelerin anlamlarını, bilişsel eş anlamlılarını, türlerini, diğer anlamlar ile arasındaki ilişkilerini ve bu anlamların tanımlarını temsil eder. Bu çalışma ile, emlak alanına özgü bir sözlük oluşturmak ve bu yeni sözlüğü kullanarak daha küçük bir kelime ağı tasarlamak yoluyla biçimbilimsel çözümleme ve anlam belirsizliği giderme gibi Doğal Dil İşleme görevlerini kolaylaştıracak bir yöntem öneriyoruz. Ön çalışma olarak, emlak alanına özgü 7,000 kelime içeren bir sözlük ve yaklaşık 11,000 eş anlam kümesinden oluşan bir kelime ağı oluşturuldu ve bunlar çeşitli görevlerle doğrulandı. WordNet is a thesaurus which represents word senses, meanings of lemmas, relations between senses, and definitions of these senses. In this paper, we propose a method to simplify the morphological analysis and disambiguation tasks performed in a specific domain by creating a subset of original Turkish dictionary and WordNet. As a preliminary study, we create a dictionary for the estate domain with 7,000 words and a WordNet with 11,000 synsets, then validate them on simple tasks. We would like to thank Mehmet Gallen for discussions. This work has been supported by the Turkish Academy of Sciences in the framework of the Young Scientist Award Program (EA-TUBA-GEBIP/2001-1-1), Bogazici University Scientific Research Project 05HA101 and Turkish Scientific Technical Research Council TUBITAK EEEAG 104E079 Publisher's Version

4th International Conference on Computer Science and Engineering, UBMK 2019 -- 11 September 2019 through 15 September 2019 -- -- 154916, 2-s2.0-85076204108, Today, with the increase of Internet-based documents, we are presented with many data that need to be processed and evaluated. Media, news and advertising are some of the areas where these data are evaluated. For the news, the classification of people in the media sector is an important problem in terms of time. In this paper, it is aimed to determine which types of news titles belong to. The dataset consists of 4200 Turkish new titles belonging to 7 class labels. In order to determine the types, classical Latent Dirichlet Allocation (LDA), Latent Semantic Analysis (LSA) and Non-Negative Matrix Factorization (NMF) algorithms were used in topic modeling. In addition, the LDA-based n-LDA method was also used. The accuracy of all methods used was measured and compared. NMF was the most successful method for three classes, while for five and seven classes LSA was the most successful method. © 2019 IEEE.

Medical Technologies Congress (TIPTEKNO) -- OCT 03-05, 2019 -- Izmir, TURKEY, WOS: 000516830900058, Parkinson's disease can be diagnosed by the speech signals. In general, the data obtained by feature extraction algorithms from the speech signals are used in any classification algorithm. Some of the extracted features have a high ability to represent the relevant problem, while others are low. In the diagnosis of Parkinson's disease, it is very important to determine which of the extracted features from the speech signals may increase the classification performance. In this paper, Artificial Bee Colony algorithm based feature selection approach is proposed for the solution of the mentioned problem. The proposed method has been analyzed in comparison with the well-known classification methods including support vector machine, k nearest neighbor, Naive Bayesian, decision tree., Biyomedikal Klinik Muhendisligi Dernegi, Izmir Katip Celebi Univ, Biyomedikal Muhendisligi Bolumu

EgeUn###, In recent studies, it is shown that word embeddings achieve in several natural language processing (NLP) tasks. Though paraphrase identification in Turkish is well-studied by traditional statistical NLP methods, to the best of our knowledge there exists no study where word and/or sentence embeddings are employed. In this paper, three methods, which are well-known as 'using average vector for word embeddings' (AWE), 'concatenated vectors for word embeddings' (CWE) and 'word mover's distance word embeddings' (WMDWE) to build sentence embeddings from word embeddings are examined and their effect in performance of paraphrase identification is measured. The results are presented comparatively for English (MSRP) and Turkish (PARDER and TuPC) paraphrase corpora. The study doesn't cover the optimization of parameters used in training of word embeddings and also the features specific to Turkish langauge are not considered. Despite this naive approach, the test results obtained from PARDER corpus are inspiring that a more detailed study that involves such improvements may result with more convincing performance values. © 2019 IEEE.

2018 International Conference on Artificial Intelligence and Data Processing, IDAP 2018 -- 28 September 2018 through 30 September 2018, Deep learning methods are successful in many different domains such as image, natural language and signal processing. However, the number of samples affects success of deep learning algorithms significantly. Therefore, it is seen as a big challenge to obtain or produce lots of labeled data. A transfer learning method has been proposed to overcome this problem. Transfer learning aimed that using a pre-trained network instead of training it from scratch as the basis for new problem. In this paper, it is looked for a solution to fabric defect detection problem through transfer learning. The sale of defective fabrics damages both producers and customers. Accurate and rapid detection of fabric defects is a crucial problem for the textile industry. Since fabric has the features of unique own textures, it is a matter of curiosity how the transfer learning method will result in determining the fabric defect. In this study, using the AlexNet model trained with millions of images, the success rate of training from stratch to 75% was increased to 98% with transfer learning. © 2018 IEEE.

4th International Conference on Computer Science and Engineering (UBMK) -- SEP 11-15, 2019 -- Samsun, TURKEY, WOS:000609879900029, Today, with the increase of Internet-based documents, we are presented with many data that need to be processed and evaluated. Media, news and advertising are some of the areas where these data arc evaluated. For the news, the classification of people in the media sector is an important problem in terms of time. in this paper, it is aimed to determine which types of news titles belong to. The dataset consists of 4200 Turkish new titles belonging to 7 class labels. in order to determine the types, classical Latent Dirichlet Allocation (LDA), Latent Semantic Analysis (LSA) and Non-Negative Matrix Factorization (NMF) algorithms were used in topic modeling. in addition, the LDA-based n-LDA method was also used. The accuracy of all methods used was measured and compared. NMF was the most successful method for three classes, while for five and seven classes LSA was the most successful method., IEEE, IEEE Turkey Sect

2018 International Conference on Artificial Intelligence and Data Processing, IDAP 2018 -- 28 September 2018 through 30 September 2018, In this paper, a realization of Sprott-94-c system using Switch State Control Cellular Neural Network (SC-CNN) is presented. In addition to simulation results, Field Programmable Gate Arrays (FPGAs) and Field Programmable Analog Arrays (FPAAs) implementations of the proposed system are also presented. Experimental results confirm the convenience of the standpoint. © 2018 IEEE.

Derin evrişimsel sinir ağları (ESA) pek çok sınıflandırma probleminde olduğu gibi anlambilimsel görüntü bölütlemede de çok ciddi başarı göstermiştir. Fakat derin ağların eğitilmesi hem zaman alıcıdır hem de geniş bir eğitim veri kümesine ihtiyaç duymaktadır. Bir veri kümesinde eğitilen ağın başka bir görev ya da veri kümesine uygulanabilmesi için transfer öğrenme ile yeniden eğitilmesi gerekmektedir. Transfer öğrenmeye alternatif olarak ağ katmanlarından çıkarılan öznitelik vektörleri doğrudan sınıflandırma amaçlı kullanılabilir. Bu bildiride genel ESA mimarilerinden elde edilen özniteliklerin eğitim gerektirmeyen bir görüntü etiketleme yönteminde kullanılmasının sınıflandırma başarımına katkısı incelenmiştir. Derin ağlarda ‘öğrenilmiş’ öznitelikler ile SIFT gibi ‘el yapımı’ özniteliklerin birlikte kullanılmasının etiketleme doğruluğunu artırdığı gösterilmiştir. Varolan ön eğitimli ağların kullanılması sayesinde önerilen yaklaşım herhangi bir veri kümesinde yeniden eğitime gerek olmadan kolayca uygulanabilmektedir. Önerilen yöntem iki veri kümesinde test edilmiş ve etiketleme doğruluğu benzer yöntemlerle karşılaştırmalı olarak sunulmuştur. Deep convolutional neural networks (CNN) have shown significant success in many classification problems including semantic image segmentation. However training of deep networks is time consuming and requires large training datasets. A network trained in one dataset could be applied to another task or dataset through transfer learning and retraining. As an alternative to transfer learning, feature vectors that are extracted from network layers could be directly used for classification purposes. In this paper we investigate the improvement in classification performance when features extracted from generic CNN architectures are used in an image labeling algorithm that does not require training. We show that the use of 'learned' features from deep networks together with 'hand-crafted' features such as SIFT increases the labeling accuracy. Since existing pre-trained networks are used, the proposed approach could be easily applied to any dataset without any retraining. The proposed method is tested in two datasets and labeling accuracies are compared with similar existing methods. Publisher's Version

Doğrudan pazarlama, uygun ürünleri uygun kişilerle en kısa yoldan buluşturma sürecidir. Son yılların en popüler pazarlama yaklaşımlarından birisidir. Bu çalışmada turizm sektörüne ait isimsizleştirilmiş bir veri tabanını kullandık. Bir otel zinciri için yapılan kampanya kapsamında veri madenciliği tekniklerini uygulayarak hedef kitle seçimi yaptık. Çalışmada birçok makine öğrenmesi yöntemini denedik. Sonuç olarak; geçmişte yapılan ve herhangi bir makine öğrenmesi yöntemi kullanılmadan hazırlanan kampanya sonuçlarına göre daha iyi sonuçlar elde ederken benzer analizlerde kullanılabilecek bir altyapı oluşturmuş olduk. Direct marketing, which is one of the most popular marketing approaches in the recent years, can be defined as the shortest and most accurate way of matching suitable products with the appropriate people. In this paper, we used an anonymous tourism sector database. We applied data mining techniques for target audience selection for a hotel campaign. As a result; we achieved better results than the results of previous campaigns without the use of any machine learning methods and we created an infrastructure that can be used for similar analyses. Publisher's Version

Bu bildiride buz patenini yeni öğrenmeye başlayanlar için hareket algılama tabanlı, ergonomik, kurulumu basit ve taşınabilir bir sistem tasarımı sunulmaktadır. Sistem kullanıcıların temel kayma tekniklerini doğru bir şekilde uygulamasına destek olacaktır. Sistem tasarlanırken, geliştirmeye açık, esnek ve modüler olmasına dikkat edilmiştir. Böylece, daha ileri seviye tekniklerin kontrolünün sağlanması veya sistemin vücut koordinasyonu gerektiren diğer sporlar için adaptasyonu daha kolay bir şekilde gerçekleştirilebilecektir. Sistemin temel destek elemanları ataletsel ölçüm birimleridir. Sistemdeki ataletsel ölçüm birimi bile¸senleri bedenin her bir uzvunun yön vektörünü belirlemektedir. Bu sensörlerden gelen veriler aracılığıyla belirlenmiş olan yönler, sistemde kullanıcıyı doğru bir şekilde yönlendirmek için altyapı oluşturmaktadır. Kullanıcı kulaklıkları ve sensör modüllerini taktıktan sonra sensörlerden gelen veriler kullanılarak kullanıcıya sesli olarak açıkça anlaşılır, motive edici ve kısa komutlar verilecektir. In this paper, an ergonomic, simple to install and portable system design is presented for those who are new to learning ice skating. The system will support users to apply basic skating techniques correctly. The system is designed to be open, flexible and modular. Therefore this system can be easily extended for more advanced techniques or other sports, requiring body coordination of the system. The main components of the system are inertial measurement units. The inertial measurement unit in the system determines the direction vector of each part of the body. The directions computed by data collected from these sensors form the basis for directing the user in the system. Once the headphones and sensor modules have been installed, the user is assisted with short, motivating and clearly understandable commands. Aselsan,et al.,Huawei,IEEE Signal Processing Society,IEEE Turkey Section,Netas Öztürk, Ö., Kahramanlı, M. M. (2018), Buz Pateni Öğrenenler İçin Hareket Algılama Tabanlı Bir Sistem Tasarımı, New York: IEEE.

Aselsan;et al.;Huawei;IEEE Signal Processing Society;IEEE Turkey Section;Netas, 26th IEEE Signal Processing and Communications Applications Conference, SIU 2018 -- 2 May 2018 through 5 May 2018 -- -- 137780, In this paper, the performance of indoor mobile multiple-input multiple-output (MIMO) visible light communication (VLC) system is investigated. Angular diversity receiver (ADR), which has a compact design, is considered at the receiver side, and the MIMO channel capacity performance is studied for different indoor scenarios. The optimum values, which provide the best MIMO channel capacity for ADR, are determined by taking into consideration the different elevation angles and distances between the light emitting diodes (LEDs). The spatial modulation (SM) bit error rate (BER) performances are obtained for ADR with optimum values and compared with the repetition coding receiver (RCR), which has low complexity and compact design. DC-informative subcarrier intensity modulation (DCI-SIM) and pulse amplitude modulation (PAM) are considered as modulation schemes. Using analytical derivations and simulation results, it is shown that the best BER performance is achieved by RCR with DCI-SIM at the center of the room. Furthermore, the BER performance of RCR with DCI-SIM is close to the performance of ADR with SM at the points near to the corners and the walls. © 2018 IEEE.

Aselsan;et al.;Huawei;IEEE Signal Processing Society;IEEE Turkey Section;Netas 26th IEEE Signal Processing and Communications Applications Conference, SIU 2018 -- 2 May 2018 through 5 May 2018 -- -- 137780 Deep convolutional neural networks (CNN) have shown significant success in many classification problems including semantic image segmentation. However training of deep networks is time consuming and requires large training datasets. A network trained in one dataset could be applied to another task or dataset through transfer learning and retraining. As an alternative to transfer learning, feature vectors that are extracted from network layers could be directly used for classification purposes. In this paper we investigate the improvement in classification performance when features extracted from generic CNN architectures are used in an image labeling algorithm that does not require training. We show that the use of 'learned' features from deep networks together with 'hand-crafted' features such as SIFT increases the labeling accuracy. Since existing pre-trained networks are used, the proposed approach could be easily applied to any dataset without any retraining. The proposed method is tested in two datasets and labeling accuracies are compared with similar existing methods. © 2018 IEEE.

Aselsan;et al.;Huawei;IEEE Signal Processing Society;IEEE Turkey Section;Netas, 26th IEEE Signal Processing and Communications Applications Conference, SIU 2018 -- 2 May 2018 through 5 May 2018 -- -- 137780, Face frontalization increases accuracies of face and gesture recognition applications. In this paper, we propose a 2D patch warping based face frontalization method which that has a simple but efficient flow due to its lower computation cost. We partition the human face into 23 nearly planar regions that are constituted by 68 landmark points to form a frontal face model and used for warping process. Planar places warped by using homography unlike other affine transform based methods. Warping rectangle regions with homography preserve global structure of face as well as it decreased the computational cost of frontalization as againts situations that work with a lot of triangular region like Delaunay triangulation. In order to test recognition performance, every test sample frontalized with respect to average face model computed as the average of all train samples. Test sets created by the pose angles of samples, tested separately to measure the contribution of proposed method to recognition and we compare the proposed method to another state of art frontalization method in literature

Aselsan;et al.;Huawei;IEEE Signal Processing Society;IEEE Turkey Section;Netas, 26th IEEE Signal Processing and Communications Applications Conference, SIU 2018 -- 2 May 2018 through 5 May 2018 -- -- 137780, NOMA is as a strong candidate for the Future Radio Access Network (FRA) due to its potential to support massive connectivity and high spectral efficiency. However, the most important drawback of NOMA is the error during Successive Interference Canceller (SIC) is implemented because of the inter-user interferences. In this paper, we derive closed-form exact Bit-Error Rate expressions for Downlink(DL) NOMA over Nakagami-m fading channels in the presence of SIC errors. The derived expressions are validated by the computer simulations. It is shown that the m parameter still represents the diversity order like as OMA systems. Besides, the BER performances of users for NOMA have substantially depended on the power allocation coefficient. © 2018 IEEE.

Publish-Subscribe(P/S) based sensor monitoring systems provide different quality-of-service according to clientserver policies. Using adaptive timing and data freshness models may increase the flexibility of services in these systems. In this paper, we propose two timing models for sensor network based emergency detection P/S Systems in which the timing constraints and data freshness can be dynamically adjusted for user needs. The performance of these models is tested in a real testbed setup. The experimental results reveal that the proposed models can meet the timing and data freshness demands of the clients for applications which acquire physical data from scalar sensors with 2-10s sampling rates. © 2018 IEEE. Aselsan;et al.;Huawei;IEEE Signal Processing Society;IEEE Turkey Section;Netas 26th IEEE Signal Processing and Communications Applications Conference, SIU 2018 -- 2 May 2018 through 5 May 2018 -- -- 137780

Aselsan;et al.;Huawei;IEEE Signal Processing Society;IEEE Turkey Section;Netas, 26th IEEE Signal Processing and Communications Applications Conference, SIU 2018 -- 2 May 2018 through 5 May 2018 -- -- 137780, Interference aligment in cognitive radio networks has recently been studied in the literature to improve the system performance. This paper investigates the performance of interference alignment based cognitive radio network under interference leakage. The closed-form mathematical expressions of the average data rate and the outage probabilty are derived under the interference leakage of the secondary network users on the primary network. The theoretical analyses for wireless communication channel is presented considering with different signal to noise ratio values under Rayleigh fading channel. © 2018 IEEE.

Aselsan;et al.;Huawei;IEEE Signal Processing Society;IEEE Turkey Section;Netas, 26th IEEE Signal Processing and Communications Applications Conference, SIU 2018 -- 2 May 2018 through 5 May 2018, This paper represents, a new heuristic algorithm, bat algorithm, is used to optimize filter coefficients in the design of low pass finite impulse response (FIR) digital filter. The FIR digital filters have a wide range of applications due to the features. Bounded input bounded output (BIBO) guarantees stability and is designed to provide linear phase at all frequencies. The bat algorithm is based on the local echolocation behaviors in the nature of the bats, and many areas such as system modeling, image processing and microwave are used. Bat algorithm was used in the design of low pass FIR digital filters for different filter degrees and the obtained results were evaluated in performance analysis. © 2018 IEEE.

Aselsan;et al.;Huawei;IEEE Signal Processing Society;IEEE Turkey Section;Netas, 26th IEEE Signal Processing and Communications Applications Conference, SIU 2018 -- 2 May 2018 through 5 May 2018 -- -- 137780, In this paper, a new method that uses radio frequency (RF) communications signal is proposed for the contactless respiration rate estimation and monitoring. In the proposed method, the phase difference between the channels of the two-channel radio receiver is used instead of the change of the RF signal amplitude with the respiration, unlike the methods in the literature. It is expected that the phase data is less affected by the environmental interferences according to the amplitude data. The subspace based MUSIC algorithm is firstly applied for the estimation of respiration rate using the change of phase difference data. The real measurements that are taken with software defined radios in laboratory, are used to validate the functionality of the proposed method. It is shown with the experimental results that the proposed MUSIC method outperforms the power spectral density based methods for the respiration rate estimation

21st National Biomedical Engineering Meeting, BIYOMUT 2017 --24 November 2017 through 26 November 2017 -- -- Diabetes is a condition where the body does not produce enough insulin to convert sugar to energy, leading to a build up of sugar in the blood. This leads to a number of problems, including diabetic retinopathy. Diabetic retinopathy is a complication of diabetes that causes damage to the blood vessels of the retina that allowing you to see fine detail. It causes progressive damage to the retina. This paper proposes a simple yet an efficient approach for automatic detection of the exudates of the Diabetic Retinopathy. The detection of exudates of diabetic retinopathy is composed of eight steps: 1. RGB to gray conversion. 2. Moving mean to 0.5 3. Max Filter 4. Threshold specification 5. Optic disk removal 6. Remove of objects which is not exudates 7. Thresholding original image using the specified exudate regions 8. Computing statistical measures. © 2017 IEEE.

Aselsan;et al.;Huawei;IEEE Signal Processing Society;IEEE Turkey Section;Netas, 26th IEEE Signal Processing and Communications Applications Conference, SIU 2018 -- 2 May 2018 through 5 May 2018 -- -- 137780, In this paper, effects of autapse on signal detection capacity of Izhikevich neuron in feed-forward-loop network motifs are investigated. Obtained results showed that autapse significantly enchances singal detection of Izhikevich neuron at proper autaptic time delay values compared without autapse. Also, it is seen that feed-forward-loop motifs have significant effects on signal detection ability of Izhikevich neuron. It is obtained that signal detection of Izhikevich neuron are best in T1 feed-forward-loop motif. © 2018 IEEE.

Aselsan;et al.;Huawei;IEEE Signal Processing Society;IEEE Turkey Section;Netas 26th IEEE Signal Processing and Communications Applications Conference, SIU 2018 -- 2 May 2018 through 5 May 2018 -- -- 137780 This paper represents efficient expectation-maximization (EM) and space-alternating generalized expectation-maximization (SAGE) algorithm based iterative channel estimation methods for multi-cell massive multiple input multiple output (MIMO) systems. The proposed iterative channel estimation methods converge to the same mean square error (MSE) performance of the least squares (LS) estimator with the increasing number of iterations. LS channel estimation method requires conjugate transpose of a large-size pilot matrix. Conjugate transpose of the large-size matrix increases computational complexity. The size of the pilot matrix depends on the number of users. As the number of users increases, computational complexity increases too. In the proposed iterative channel estimation methods, the size of the pilot matrix is reduced to the vector size and the computational complexity is significantly reduced. The SAGE algorithm can estimate the channel by performing fewer iterations than EM, so it is a preferable method in terms of convergence speed. © 2018 IEEE. Bu tez’de çok hücreli masif çok girdili çok çıktılı(MIMO) sistemler için beklenti enbüyükleme (EM) ve uzay dönü- ̧sümlü genelle ̧stirilmi ̧s beklenti enbüyükleme (SAGE) yinelemelikanal kestirim yöntemleri ele alınmı ̧stır. Bu yöntemler kanalıyinelemeli olarak kestirir ve belli sayıda yinelemeden sonra enküçük kareler (LS) kanal kestirim yöntemi ile aynı ortalamakaresel hata (MSE) performansını gösterirler. Sistemdeki kul-lanıcı sayısı oldukça fazla oldu ̆gu için LS kanal kestirimindekullanılan pilot matrisinin boyutu da oldukça büyüktür. Kanalkestirimi sırasında pilot matrisinin e ̧slenik transpozunu almak ge-rekti ̆gi için hesaplama karma ̧sıklı ̆gı oldukça yüksektir. Sistemdekikullanıcı sayısı arttıkça hesaplama karma ̧sıklı ̆gı da artmaktadır.Önerilen yinelemeli kanal kestirim yöntemlerinde pilot matrisivektör boyutuna indirgenir ve hesaplama karma ̧sıklı ̆gı önemliölçüde azaltılır. SAGE algoritması EM algoritmasına kıyasla dahaaz sayıda yineleme yaparak kanalı kestirebilmektedir, dolayısıylayakınsama hızı açısından daha tercih edilebilir bir yöntemdir.

Aselsan;et al.;Huawei;IEEE Signal Processing Society;IEEE Turkey Section;Netas, 26th IEEE Signal Processing and Communications Applications Conference, SIU 2018 -- 2 May 2018 through 5 May 2018 -- -- 137780, Automatic gender recognition has been applied in many areas such as analyzing the structure of demographic social environments. In this paper we present a gender recognition algorithm based on uniform local binary pattern (LBP). LBP has proven to be very efficient image descriptor in many pattern recognition applications. In feature extraction step of the proposed method, a well known variant of LBP, uniform local binary pattern is utilized. Before applying LBP the face detection and alignment tasks are performed respectively. By doing alignment, the overlapping of similar part of face has been achieved. Then, we combine local uniform LBP histograms to create the final feature vector. We show that in the experiments carried out in Feret face dataset that the proposed method can achieve up to 93.57% recognition rate

Bu bildiride, birden fazla teknik kullanılarak IEEE 802.11af ve IEEE 802.22 kablosuz sistemlerinin TV beyaz bo¸slugunda birlikte varolabilmeleri için algoritmalar geli¸stirilmi¸stir. ˘Me¸sgul tonu, güç kontrolü ve çok sekmeli haberle¸sme teknikleri kullanılarak daha önce yapılan çalı¸smalar temelinde iyile¸stirmeler yapılmı¸stır. IEEE 802.22 sistemlerinin korunumu birinci planda yer alırken, aynı zamanda IEEE 802.11af sistemlerininde haberle¸smeye devam etmesi saglanmı¸stır. Standartlara uygun ˘haberle¸sme parametreleri ve kanal modelleri kullanılan bu çalı¸smada, her iki sistemin de paket giri¸sim oranları ve ba¸sarılı paket oranları farklı senaryolar için incelenmi¸s ve iyile¸stirilmi¸s sonuçlar ortaya konulmu¸stur. In this paper, various algorithms are proposed to help IEEE 802.11af and IEEE 802.22 wireless systems to coexist in TV white space by considering multiple techniques. By using busy tone, power control and multihop communication techniques, improvements have been made compared to prior approaches. With this approach, while IEEE 802.22 systems are protected from intra-network interference, IEEE 802.11af systems are still able to continue transmitting packets. Considering the communication parameters and channel models compliant for the standards, IEEE 802.11af and IEEE 802.22 system performances are investigated in terms of interfering packet rate and successful packet transmission rate for different scenarios and the improved results are presented. IEEE Huawei Aselsan NETAS

25th Signal Processing and Communications Applications Conference, SIU 2017 -- 15 May 2017 through 18 May 2017 -- -- 128703 Large scale dictionaries such as wordnets are one of the important structures used in natural language processing. Wordnet is a comprehensive dictionary containing semantic relations including synonyms, antonyms, etc. In this paper, we try to extract Hypernym-Hyponym relations which are one of the main parts of WordNet. For this aim, we used Wikipedia, Contemporary Dictionary of Turkish and Wiktionary as corpora. We used Finite State Machines for developing several rules for Hypernym-Hyponym extraction. © 2017 IEEE.

25th Signal Processing and Communications Applications Conference, SIU 2017 -- 15 May 2017 through 18 May 2017 -- -- 128703, Reverberation and multipath propagation afîect the accuracy of time of flight (TOF) estimation in confined underwater medium. In this paper, time delay has been estimated by obtaining the transmitted signal and first echo signal from measured signal acquised by using a wideband transducer with 5 MHz central frequency. Cross-correlation (CC) and generalized cross-correlation (GCC) methods have been applied to the obtained signals and their performances have been compared. The results show that GCC method has better performance for time delay estimation in confined underwater medium where multipath propagation and reverberation are severe. © 2017 IEEE.

Bu bildiride, IEEE 802.22 ve IEEE 802.11af kablosuz sistemlerinin TV beyaz boşluğunda uyumlu çalışabilmeleri için meşgul ton tabanlı yeni bir güç kontrol algoritması önerilmiştir. Daha önceki meşgul ton tabanlı çalışmalardan farklı olarak, 802.11af erişim noktası ve erişim noktasına bağlı kullanıcılar meşgul ton sinyalini algıladıktan sonra haberleşme güçlerini konum bilgilerine göre ayarlamakta ve gerekiyorsa iki hop üzerinden haberleşmektedirler. Böylelikle 802.22 sistemlerine yapılan girişim azaltılırken, 802.11af sistemlerinin haberleşmesi de mümkün kılınmaktadır. Bu çalışmada standartlara uygun haberleşme parametreleri ve kanal modelleri dikkate alınarak, 802.11af ve 802.22 sistemlerinin performansları paket girişim oranı ve başarılı paket oranı türlerinden farklı senaryolar altında incelenmiştir. In this paper, a new power control algorithm based on busy tone approach has been proposed for the coordination of IEEE 802.22 and IEEE 802.11af systems in TV white space. Different from the earlier studies, in addition to both 802.11af access point and clients listening to the busy tone, they also adjust their communication power according to the location information and use hopping for communication, if needed. Acccordingly, interference caused to 802.22 systems has been reduced while the 802.11af systems are still able to communicate. This study quantifies the 802.11af and 802.22 system performances in terms of interfering packet rate and succesful packet transmission rate for different scenarios considering the communication parameters and channel models adapted for the standards.

2016 Medical Technologies National Conference, TIPTEKNO 2016 -- 27 October 2016 through 29 October 2016 -- -- 126633, In this paper; the effects of autapse (a kind of synapse formed between the axon of the soma of a neuron and its own dendrites) and ion channel blocking on the firing regularity of a biological small-world neuronal network, consists of stochastic Hodgkin-Huxley neurons, are studied. In this study, it is assumed that all of the neurons on the network have a chemical autapse and constant membrane area. Obtained results indicate that there are different effects of channel blockage and parameters of the autapse on the regularity of the network. It is found that the firing regularity of the network is decreased with the sodium channel blockage while increased with potassium channel blockage. Besides, it is determined that regularity of the network augments with the conductance of the autapse. © 2016 IEEE.

2016 Medical Technologies National Conference, TIPTEKNO 2016 -- 27 October 2016 through 29 October 2016 -- -- 126633, Sensing and measurement methods with Quartz Crystal Microbalance (QCM) based sensors draw attention as an improving application among studies in recent years. This study is carried out for to be basis of studies on glucose level measurement using human blood serum using QCM sensors. In this paper, measurements with QCM sensor system using pure water containing several levels of glucose samples are carried out. Measurement results are analyzed with plots. The obtained results show that QCM based sensors are sensitive, repeatable and reliable sensor applications for blood glucose measurements. © 2016 IEEE.

2016 Medical Technologies National Conference, TIPTEKNO 2016 -- 27 October 2016 through 29 October 2016 -- -- 126633, In this paper, the effects of autapse (a kind of synapse formed between the axon or soma of a neuron and its own dendrite) on the transmission of weak signal are investigated in scale-free neuronal networks. In the study, we consider that each neuron in the network has an autapse modelled as chemical synapse. Besides, a subthreshold signal are injected to all neurons in the network. When obtained results are analyzed, it is seen that the autaptic conductance has a threshold value. According to this threshold, the effects of autapse on the transmission of weak signal are evaluated within two categories. When conductance of autapse is less than this threshold value, autapses with proper autaptic time delay values slightly increase the transmission of weak signal. When autaptic conductance is bigger than the threshold value, autapses with proper autaptic time deley values prominently block the transmission of weak signal. © 2016 IEEE.

25th Signal Processing and Communications Applications Conference, SIU 2017 --15 May 2017 through 18 May 2017 -- -- The Internet and social media provide a major source of information about people's opinions. Due to the rapidly growing number of online documents, it becomes both time-consuming and hard task to obtain and analyze the desired opinionated information. Sentiment analysis is the classification of sentiments expressed in documents. To improve classification perfromance feature selection methods which help to identify the most valuable features are generally applied. In this paper, we compare the performance of four feature selection methods namely Chi-square, Information Gain, Query Expansion Ranking, and Ant Colony Optimization using Maximum Entropi Modeling classification algorithm over Turkish Twitter dataset. Therefore, the effects of feature selection methods over the performance of sentiment analysis of Turkish Twitter data are evaluated. Experimental results show that Query Expansion Ranking and Ant Colony Optimization methods outperform other traditional feature selection methods for sentiment analysis. © 2017 IEEE.

25th Signal Processing and Communications Applications Conference, SIU 2017 -- 15 May 2017 through 18 May 2017 -- -- 128703, In this paper, a study which automatically detects four open string notes of the violin from the recorded violin sounds is presented for the purpose of using in the beginner level violin education. First, in order to be able to analyze the violin sound recordings, it is necessary for the music piece to be segmented into the notes forming the music piece itself. For this, it is necessary to correctly determine the starting times (onsets) of the notes. Because of the envelope characteristic that the violin sound has, it is more difficult to detect the note onset than other musical instrument sounds such as piano and guitar. By using the complex spectral difference method, the note onsets were determined, then by determining the fundamental frequency the note of the violin sound was found. The proposed system was applied to the open string sound tracks obtained from eight people by using the iPad-based professional recording system, and the results showed that the proposed system worked successfully. © 2017 IEEE.

2016 National Conference on Electrical, Electronics and Biomedical Engineering, ELECO 2016 --1 December 2016 through 3 December 2016 -- -- Nowadays, magnetically controlled shunt reactors (MCSRs)have been widely used to solve the problems of the power quality. MCSR can simplify the system reactive power and voltage control in the super/special high-voltage power grid, suppress power frequency and operation over voltage, eliminating generator self excitation, dynamically compensate charging the power in a transmission line, suppress the secondary arc current, damping system resonance and so on. This paper presents a comprehensive review of MCSR control strategies. It is aimed at providing a broad perspective on the status of MCSR technology to the researchers and the application engineers dealing with power quality issues. A list of more than 50 research publications on the subject is also appended for a quick reference. © 2016 The Chamber of Turkish Electrical Engineers.

Medical Technologies National Congress (TIPTEKNO) -- OCT 12-14, 2017 -- TRABZON, TURKEY, WOS: 000427649500045, In this paper, the effects of autapse (a kind of synapse formed between the axon or soma of a neuron and its own dendrites) on the weak signal detection capacity of a Hodgkin-Huxley (H-H) neuron are investigated. In the study, we consider that the H-H neuron has an inhibitory autapse modeled as a chemical synapse. The subthreshold sine wave is injected to the H-H neuron as a weak signal. Obtained results indicate that inhibitory autapse prominently increases the weak signal detection capacity of a H-H neuron when the proper autaptic time delay and autaptic conductance values are choosen., IEEE Turkey Sect

25th Signal Processing and Communications Applications Conference, SIU 2017 -- 15 May 2017 through 18 May 2017 -- -- 128703, In this paper, a spectral graph based image coding method is proposed. In the proposed method, only the smoothing parameter in the graph adjacency matrix, graph Fourier transform (GFT) coefficients kept after hard thresholding and column indices of eigenvectors corresponding to kept coefficients are sent to the decoder side. It is advantageous in that this coding process is performed by sending much less information as compared to other graph based methods. Our results indicate that our proposed method enables image compression without deteriorating the visual quality by using very small numbers of GFT coefficients

25th Signal Processing and Communications Applications Conference, SIU 2017 -- 15 May 2017 through 18 May 2017 -- -- 128703, In this paper, the performance of direct current informative subcarrier intensity modulation (DCI-SIM) which is proposed to enhance the power efficiency of subcarrier intensity modulation (SIM) is investigated in visible light communication. The BER performance of DCI-SIM is compared with SIM, pulse amplitude modulation (PAM) and, enhanced unipolar orthogonal frequency division multiplexing (eU-OFDM) for different spectral efficiencies. Using analytical derivations and simulation results, we show that the DCI-SIM is the most power efficient modulation technique among others, and that the power efficiency increases with the spectral efficiency. Furthermore, we optimize the constellation of 64-DCI-SIM and present the results. © 2017 IEEE.

24th Signal Processing and Communication Application Conference, SIU 2016 -- 16 May 2016 through 19 May 2016 -- -- 122605, In this paper, 11 different power quality disturbances were automatically detected by using statistical features with wavelet transform and norm entropy techniques. The best of the created features were selected with forward selection algorithm. Performance of classification algorithms, Support Vector Machines (SVM), Multi Layer Perceptron (MLP), k Nearest Neighbor (KNN) and random subspace KNN (Sub-KNN) which is an ensemble method, were examined. Consequently, the best classification accuracy of 99.3% was achieved by using Sub-KNN and it was appeared that compared to other methods, this algorithm was more robust against the noise. © 2016 IEEE.

24th Signal Processing and Communication Application Conference, SIU 2016 -- 16 May 2016 through 19 May 2016 -- -- 122605 Medical images captured by various devices have different illumination states based on chemicals used by patient prior to scanning. Consider a MRI image which has low contrast or is too bright, hence the experts cannot analysis that image due to poor representation of data in the image. In this paper we are proposing new medical image illumination enhancement and sharpening technique based on stationary wavelet transform which is addressing the aforementioned problem. The technique decomposes the input medical image into the four frequency subbands by using stationary wavelet transformation and enhances the illumination of the low-low subband image, and then it enhanced edges of image by adding the high frequency subbands to the image. The technique is compared with the conventional and state-of-art image illumination enhancement techniques such as histogram equalisation, local histogram equalisation, singular value equalisation, and discrete wavelet transform followed by singular value decomposition contrast enhancement techniques. The experimental results are showing the superiority of the proposed method over the conventional and the state-of-art techniques. © 2016 IEEE.

In this paper a new computationally efficient and high performance channel estimation algorithm is proposed for the indoor visible light communication (VLC) sparse channels in the presence of a clipping noise. The clipping noise is modelled as a Gaussian mixture and a first time in the literature the matching pursuit (MP) and the space-alternating expectation- maximization (SAGE) algorithms are combined into the new estimation called the SAGE-MP algorithm for iteratively estimating the sparse channel coefficients as well as their positions efficiently. The MP algorithm is also employed to determine the initial values of the joint iterative algorithm. Computed simulations indicates that the SAGE-MP algorithm converge in 3 iterations at most and yields excellent bit error rate (BER) and mean-square error (MSE) performances for DC-biased optical OFDM (DCO-OFDM) based systems. © 2016 IEEE.

In this paper a new optical orthogonal frequency division multiplexing (OFDM) method is proposed for multiple-input multiple-output (MIMO) visible light communications (VLC) systems. In the new scheme non-DC biased optical OFDM (NDC-OFDM) and optical spatial modulation (OSM) methods are combined to obtain a novel optical OFDM scheme which has a better error performance. The new scheme is named as non-DC biased optical spatial modulation OFDM (NDC-OSM-OFDM). It is shown via computer simulations that the proposed scheme achieves better bit error rate (BER) performance than DC-biased optical spatial modulation OFDM (OSM-OFDM-DC) asymmetrically clipped optical spatial modulation OFDM (OSM-OFDM-AC) and NDC-OFDM. © 2016 IEEE.