7 results
Search Results
Now showing 1 - 7 of 7
Article Citation - WoS: 1Citation - Scopus: 1Machine Vs. Deep Learning Comparision for Developing an International Sign Language Translator(Taylor & Francis Ltd, 2022) Eryilmaz, Meltem; Balkaya, Ecem; Ucan, Eylul; Turan, Gizem; Oral, Seden GulayThis study aims to enable deaf and hard-of-hearing people to communicate with other individuals who know and do not know sign language. The mobile application was developed for video classification by using MediaPipe Library in the study. While doing this, considering the problems that deaf and hearing loss individuals face in Turkey and abroad modelling and training stages were carried out with the English language option. With the real-time translation feature added to the study individuals were provided with instant communication. In this way, communication problems experienced by hearing-impaired individuals will be greatly reduced. Machine learning and Deep learning concepts were investigated in the study. Model creation and training stages were carried out using VGG16, OpenCV, Pandas, Keras, and Os libraries. Due to the low success rate in the model created using VGG16, the MediaPipe library was used in the formation and training stages of the model. The reason for this is that, thanks to the solutions available in the MediaPipe library, it can normalise the coordinates in 3D by marking the regions to be detected in the human body. Being able to extract the coordinates independently of the background and body type in the videos in the dataset increases the success rate of the model in the formation and training stages. As a result of an experiment, the accuracy rate of the deep learning model is 85% and the application can be easily integrated with different languages. It is concluded that deep learning model is more accure than machine learning one and the communication problem faced by hearing-impaired individuals in many countries can be reduced easily.Article Citation - WoS: 29Citation - Scopus: 43Text Classification Using Improved Bidirectional Transformer(Wiley, 2022) Tezgider, Murat; Yıldız, Beytullah; Yildiz, Beytullah; Aydin, Galip; Yıldız, BeytullahText data have an important place in our daily life. A huge amount of text data is generated everyday. As a result, automation becomes necessary to handle these large text data. Recently, we are witnessing important developments with the adaptation of new approaches in text processing. Attention mechanisms and transformers are emerging as methods with significant potential for text processing. In this study, we introduced a bidirectional transformer (BiTransformer) constructed using two transformer encoder blocks that utilize bidirectional position encoding to take into account the forward and backward position information of text data. We also created models to evaluate the contribution of attention mechanisms to the classification process. Four models, including long short term memory, attention, transformer, and BiTransformer, were used to conduct experiments on a large Turkish text dataset consisting of 30 categories. The effect of using pretrained embedding on models was also investigated. Experimental results show that the classification models using transformer and attention give promising results compared with classical deep learning models. We observed that the BiTransformer we proposed showed superior performance in text classification.Article Citation - WoS: 103Citation - Scopus: 160Cassava Disease Recognition From Low-Quality Images Using Enhanced Data Augmentation Model and Deep Learning(Wiley, 2021) Abayomi-Alli, Olusola Oluwakemi; Damasevicius, Robertas; Misra, Sanjay; Maskeliunas, RytisImprovement of deep learning algorithms in smart agriculture is important to support the early detection of plant diseases, thereby improving crop yields. Data acquisition for machine learning applications is an expensive task due to the requirements of expert knowledge and professional equipment. The usability of any application in a real-world setting is often limited by unskilled users and the limitations of devices used for acquiring images for classification. We aim to improve the accuracy of deep learning models on low-quality test images using data augmentation techniques for neural network training. We generate synthetic images with a modified colour value distribution to expand the trainable image colour space and to train the neural network to recognize important colour-based features, which are less sensitive to the deficiencies of low-quality images such as those affected by blurring or motion. This paper introduces a novel image colour histogram transformation technique for generating synthetic images for data augmentation in image classification tasks. The approach is based on the convolution of the Chebyshev orthogonal functions with the probability distribution functions of image colour histograms. To validate our proposed model, we used four methods (resolution down-sampling, Gaussian blurring, motion blur, and overexposure) for reducing image quality from the Cassava leaf disease dataset. The results based on the modified MobileNetV2 neural network showed a statistically significant improvement of cassava leaf disease recognition accuracy on lower-quality testing images when compared with the baseline network. The model can be easily deployed for recognizing and detecting cassava leaf diseases in lower quality images, which is a major factor in practical data acquisition.Conference Object Citation - Scopus: 1Toxicity Detection Using State of the Art Natural Language Methodologies(Ieee, 2023) Keskin, Enes Faruk; Acikgoz, Erkut; Dogan, GulustanIn this paper, the studies carried out to detect objectionable expressions in any text will be explained. Experiments were performed with Sentence transformers, supervised machine learning algorithms, and Bert transformer architecture trained in English, and the results were observed. To prepare the dataset used in the experiments, the natural language processing and machine learning methodologies of the toxic and non-toxic contents in the labeled text data obtained from the Kaggle platform are explained, and then the methods and performances of the models trained using this dataset are summarized in this paper.Conference Object Citation - Scopus: 6A Mini-Review on Radio Frequency Fingerprinting Localization in Outdoor Environments: Recent Advances and Challenges(Institute of Electrical and Electronics Engineers Inc., 2022) Dogan,D.; Dalveren,Y.; Kara,A.A considerable growth in demand for locating the source of emissions in outdoor environments has led to the rapid development of various localization methods. Among these, RF fingerprinting (RFF) localization has become one of the most promising method due to its unique advantages resulted from the recent developments in machine learning techniques. In this short review, it is aimed to assess the existing RFF methods in the literature for outdoor localization. For this purpose, firstly, the current state of RFF localization methods in outdoor environments are overviewed. Then, the main research challenges in the development of RFF localization are highlighted. This is followed by a brief discussion on the open issues in order to give future research directions. Furthermore, the research efforts currently undertaken by the authors are briefly addressed. © 2022 IEEE.Article Citation - WoS: 6Citation - Scopus: 10Beyond Rouge: a Comprehensive Evaluation Metric for Abstractive Summarization Leveraging Similarity, Entailment, and Acceptability(World Scientific Publ Co Pte Ltd, 2024) Briman, Mohammed Khalid Hilmi; Yıldız, Beytullah; Yildiz, Beytullah; Yıldız, BeytullahA vast amount of textual information on the internet has amplified the importance of text summarization models. Abstractive summarization generates original words and sentences that may not exist in the source document to be summarized. Such abstractive models may suffer from shortcomings such as linguistic acceptability and hallucinations. Recall-Oriented Understudy for Gisting Evaluation (ROUGE) is a metric commonly used to evaluate abstractive summarization models. However, due to its n-gram-based approach, it ignores several critical linguistic aspects. In this work, we propose Similarity, Entailment, and Acceptability Score (SEAScore), an automatic evaluation metric for evaluating abstractive text summarization models using the power of state-of-the-art pre-trained language models. SEAScore comprises three language models (LMs) that extract meaningful linguistic features from candidate and reference summaries and a weighted sum aggregator that computes an evaluation score. Experimental results show that our LM-based SEAScore metric correlates better with human judgment than standard evaluation metrics such as ROUGE-N and BERTScore.Article Citation - WoS: 9Citation - Scopus: 13Improving Word Embedding Quality With Innovative Automated Approaches To Hyperparameters(Wiley, 2021) Yildiz, Beytullah; Yıldız, Beytullah; Tezgider, Murat; Yıldız, BeytullahDeep learning practices have a great impact in many areas. Big data and significant hardware developments are the main reasons behind deep learning success. Recent advances in deep learning have led to significant improvements in text analysis and classification. Progress in the quality of word representation is an important factor among these improvements. In this study, we aimed to develop word2vec word representation, also called embedding, by automatically optimizing hyperparameters. Minimum word count, vector size, window size, negative sample, and iteration number were used to improve word embedding. We introduce two approaches for setting hyperparameters that are faster than grid search and random search. Word embeddings were created using documents of approximately 300 million words. We measured the quality of word embedding using a deep learning classification model on documents of 10 different classes. It was observed that the optimization of the values of hyperparameters alone increased classification success by 9%. In addition, we demonstrate the benefits of our approaches by comparing the semantic and syntactic relations between word embedding using default and optimized hyperparameters.
