Feyza Cerezci̇
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Feyza Cerezci̇ has a total of 20 co-authors. Here are the top co-authors by number of shared publications:
- Serap Çakar Kaman
Sakarya University
4 shared publications - Gülüzar Çit
Sakarya Üniversitesi
3 shared publications - Ahmet Furkan Sönmez
Zonguldak Bülent Ecevit University
2 shared publications - Cemil Öz
Sakarya University
2 shared publications - Firgat Muradli̇
SAKARYA ÜNİVERSİTES
2 shared publications - Muhammed Kotan
Sakarya University
2 shared publications
Latest Publications
- 31 December 2024
- Published by Turkish Journal of Mathematics and Computer Science, Association of Mathematicians in Turkish Journal of Mathematics and Computer Science
- Vol. 16 (2) , 325-332
Mask-Based IoU for Bounding Box Regression Using Medical Images
Bounding box regression plays a pivotal role in the majority of object detection algorithms, significantly influencing the accuracy of object positioning and the regression speed of Convolutional Neural Networks (CNN). In object detection benchmarks, Intersection over Union (IoU) remains the widely adopted metric for evaluation. Traditional IoU-based loss functions often suffer from poor training outcomes and slow convergence, and they fail to account for situations where the predicted bounding box does not entirely capture the object’s mask. This study introduces the Mask-based Intersection over Union (MbIoU) metric for improving bounding box regression in object detection using medical images. The proposed MbIoU metric incorporates the object mask into the bounding box regression process, offering a more precise evaluation of how well the predicted bounding box encapsulates the object. The developed MbIoU metric was tested on the MNIST: HAM10000 dermoscopic skin images dataset, COVID-19 CT dataset, and Brain Tumor dataset and compared to traditional IoU metrics. The results show that MbIoU enhances the prediction by better capturing the object’s contained mask.
- 31 December 2024
- Published by Turkish Journal of Mathematics and Computer Science, Association of Mathematicians in Turkish Journal of Mathematics and Computer Science
- Vol. 16 (2) , 325-332
Mask-Based IoU for Bounding Box Regression Using Medical Images
Bounding box regression plays a pivotal role in the majority of object detection algorithms, significantly influencing the accuracy of object positioning and the regression speed of Convolutional Neural Networks (CNN). In object detection benchmarks, Intersection over Union (IoU) remains the widely adopted metric for evaluation. Traditional IoU-based loss functions often suffer from poor training outcomes and slow convergence, and they fail to account for situations where the predicted bounding box does not entirely capture the object’s mask. This study introduces the Mask-based Intersection over Union (MbIoU) metric for improving bounding box regression in object detection using medical images. The proposed MbIoU metric incorporates the object mask into the bounding box regression process, offering a more precise evaluation of how well the predicted bounding box encapsulates the object. The developed MbIoU metric was tested on the MNIST: HAM10000 dermoscopic skin images dataset, COVID-19 CT dataset, and Brain Tumor dataset and compared to traditional IoU metrics. The results show that MbIoU enhances the prediction by better capturing the object’s contained mask.
- Open Access
- 26 January 2024
- Published by Duzce Universitesi Bilim ve Teknoloji Dergisi in Düzce Üniversitesi Bilim ve Teknoloji Dergisi
- Vol. 12 (1) , 247-263
Elektromanyetik Radyasyon Maruziyet Algısı Ölçeğinin Geliştirilmesi: Geçerlilik ve Güvenilirlik Analizleri
Teknolojinin ilerlemesiyle birlikte yeni çevre sorunlarıyla karşılaşmaktayız. Son zamanlarda, duyularımızla fark edemediğimiz, iletişim teknolojilerinin ve elektriğin yaygın kullanımının çevresel bir atığı olan elektromanyetik radyasyonun olumsuz sağlık sonuçları tartışılmaktadır. Elektromanyetik radyasyon seviyesinin oluşturduğu olumsuz etkiler ile alınacak kişisel ve toplumsal önlemler için algı çalışmalarının yapılması gereklilik haline gelmiştir. Bu çalışmada, elektromanyetik radyasyon maruziyet algısı ile önlem almaya yönelik algı seviyesinin boyutsal olarak ölçülebilmesi amacıyla 1187 kişiden oluşan bir çalışma grubunda Elektromanyetik Radyasyon Maruziyet Algısı Ölçeğinin geliştirilmesi amaçlanmıştır. Yapılan geçerlilik ve güvenilirlik çalışmaları sonucunda, 11 madde ve 3 boyuttan (Farkındalık, Bilinçlilik, Endişe) oluşan geçerli ve güvenilir bir ölçek elde edilmiştir. Ölçek toplam varyansın %57,6’sını açıklamıştır. Ayrıca yapılan doğrulayıcı faktör analizi ile elde edilen üç faktörlü ölçek yapısı modellenmiş ve doğrulanmıştır. Elde edilen uyum iyiliği değerleri kabul edilebilir aralıkta bulunmuştur (Ki-kare/serbestlik değeri: 2,94; RMSEA: ,07; NFI: ,91; NNFI: ,92; CFI: ,96; GFI: ,92; AGFI: ,87). Ölçeğin güvenilirlik katsayısı Cronbach Alpha değeri “Farkındalık” boyutu için ,82, “Bilinçlilik” boyutu için ,74, “Endişe” boyutu için ,64 ve ölçek toplamı için ,77 olarak bulunmuştur. Böylece, Elektromanyetik Radyasyon Maruziyet Algısı Ölçeği adında geçerli ve güvenilir bir ölçek ortaya çıkmıştır.
- 26 January 2024
- Published by Duzce Universitesi Bilim ve Teknoloji Dergisi in Düzce Üniversitesi Bilim ve Teknoloji Dergisi
- Vol. 12 (1) , 247-263
Open AccessElektromanyetik Radyasyon Maruziyet Algısı Ölçeğinin Geliştirilmesi: Geçerlilik ve Güvenilirlik Analizleri
Teknolojinin ilerlemesiyle birlikte yeni çevre sorunlarıyla karşılaşmaktayız. Son zamanlarda, duyularımızla fark edemediğimiz, iletişim teknolojilerinin ve elektriğin yaygın kullanımının çevresel bir atığı olan elektromanyetik radyasyonun olumsuz sağlık sonuçları tartışılmaktadır. Elektromanyetik radyasyon seviyesinin oluşturduğu olumsuz etkiler ile alınacak kişisel ve toplumsal önlemler için algı çalışmalarının yapılması gereklilik haline gelmiştir. Bu çalışmada, elektromanyetik radyasyon maruziyet algısı ile önlem almaya yönelik algı seviyesinin boyutsal olarak ölçülebilmesi amacıyla 1187 kişiden oluşan bir çalışma grubunda Elektromanyetik Radyasyon Maruziyet Algısı Ölçeğinin geliştirilmesi amaçlanmıştır. Yapılan geçerlilik ve güvenilirlik çalışmaları sonucunda, 11 madde ve 3 boyuttan (Farkındalık, Bilinçlilik, Endişe) oluşan geçerli ve güvenilir bir ölçek elde edilmiştir. Ölçek toplam varyansın %57,6’sını açıklamıştır. Ayrıca yapılan doğrulayıcı faktör analizi ile elde edilen üç faktörlü ölçek yapısı modellenmiş ve doğrulanmıştır. Elde edilen uyum iyiliği değerleri kabul edilebilir aralıkta bulunmuştur (Ki-kare/serbestlik değeri: 2,94; RMSEA: ,07; NFI: ,91; NNFI: ,92; CFI: ,96; GFI: ,92; AGFI: ,87). Ölçeğin güvenilirlik katsayısı Cronbach Alpha değeri “Farkındalık” boyutu için ,82, “Bilinçlilik” boyutu için ,74, “Endişe” boyutu için ,64 ve ölçek toplamı için ,77 olarak bulunmuştur. Böylece, Elektromanyetik Radyasyon Maruziyet Algısı Ölçeği adında geçerli ve güvenilir bir ölçek ortaya çıkmıştır.
- 18 October 2023
- Published by Sakarya University Journal of Science in Sakarya University Journal of Science
- Vol. 27 (5) , 1079-1087
Estimating Human Poses Using Deep Learning Model
Over the past decade, extensive research has focused on the extraction of 3D human poses from images. The existing datasets must effectively address common challenges related to pose estimation. These datasets serve as valuable resources for evaluating, informing, and comparing different models. Deep learning models have gained widespread adoption and have demonstrated impressive performance across various domains of research and engineering. In this study, we employ these models, leveraging the open-source libraries OpenCV and Keras. To enhance the diversity and complexity of the training and testing process, we utilize the MPII Human Pose dataset. Specifically, we train and test the ResNet50 and VGG16 models using this dataset, resulting in significant improvements. The model's performance is evaluated based on the validation rate of the dataset and the accuracy of our model was 88.8 percent for VGG16 and 67 percent for ResNet50.
- 18 October 2023
- Published by Sakarya University Journal of Science in Sakarya University Journal of Science
- Vol. 27 (5) , 1079-1087
Estimating Human Poses Using Deep Learning Model
Over the past decade, extensive research has focused on the extraction of 3D human poses from images. The existing datasets must effectively address common challenges related to pose estimation. These datasets serve as valuable resources for evaluating, informing, and comparing different models. Deep learning models have gained widespread adoption and have demonstrated impressive performance across various domains of research and engineering. In this study, we employ these models, leveraging the open-source libraries OpenCV and Keras. To enhance the diversity and complexity of the training and testing process, we utilize the MPII Human Pose dataset. Specifically, we train and test the ResNet50 and VGG16 models using this dataset, resulting in significant improvements. The model's performance is evaluated based on the validation rate of the dataset and the accuracy of our model was 88.8 percent for VGG16 and 67 percent for ResNet50.
- 31 August 2023
- Published by Sakarya University Journal of Computer and Information Sciences in Sakarya University Journal of Computer and Information Sciences
- Vol. 6 (2) , 114-122
Deep Learning-Based Classification of Dermoscopic Images for Skin Lesions
Skin cancer has emerged as a grave health concern leading to significant mortality rates. Diagnosis of this disease traditionally relies on specialist dermatologists who interpret dermoscopy images using the ABCD rule. However, the integration of computer-aided diagnosis technologies is gaining popularity as a means to assist clinicians in accurate skin cancer diagnosis, overcoming potential challenges associated with human error. The objective of this research is to develop a robust system for the detection of skin cancer by employing machine learning algorithms for skin lesion classification and detection. The proposed system utilizes Convolutional Neural Network (CNN), a highly accurate and efficient deep learning technique well-suited for image classification tasks. By using the power of CNN, this system effectively classifies various skin diseases in dermoscopic images associated with skin cancer The MNIST HAM10000 dataset, comprising 10015 images, serves as the foundation for this study. The dataset encompasses seven distinct skin diseases falling within the realm of skin cancer. In this study, diverse transfer learning methods were used and evaluated to enhance the performance of the system. By comparing and analyzing these approaches the highest accuracy rate was obtained using the MobileNetV2 model with a rate of 80.79% accuracy.
- 31 August 2023
- Published by Sakarya University Journal of Computer and Information Sciences in Sakarya University Journal of Computer and Information Sciences
- Vol. 6 (2) , 114-122
Deep Learning-Based Classification of Dermoscopic Images for Skin Lesions
Skin cancer has emerged as a grave health concern leading to significant mortality rates. Diagnosis of this disease traditionally relies on specialist dermatologists who interpret dermoscopy images using the ABCD rule. However, the integration of computer-aided diagnosis technologies is gaining popularity as a means to assist clinicians in accurate skin cancer diagnosis, overcoming potential challenges associated with human error. The objective of this research is to develop a robust system for the detection of skin cancer by employing machine learning algorithms for skin lesion classification and detection. The proposed system utilizes Convolutional Neural Network (CNN), a highly accurate and efficient deep learning technique well-suited for image classification tasks. By using the power of CNN, this system effectively classifies various skin diseases in dermoscopic images associated with skin cancer The MNIST HAM10000 dataset, comprising 10015 images, serves as the foundation for this study. The dataset encompasses seven distinct skin diseases falling within the realm of skin cancer. In this study, diverse transfer learning methods were used and evaluated to enhance the performance of the system. By comparing and analyzing these approaches the highest accuracy rate was obtained using the MobileNetV2 model with a rate of 80.79% accuracy.
- 3 March 2023
- Published by Sakarya University Journal of Science in Sakarya University Journal of Science
Human Pose Estimation Using Deep Learning Models
In the last decade, there has been a lot of research into recovering 3D human poses from images. The current datasets need to address frequent exposure estimation problems adequately. However, these are shared resources to assess, inform, and contrast various models. Deep learning models are widely utilized and perform at high levels in many branches of research and engineering. So these models are used in this study with the help of the OpenCV and Keras libraries, which are open-source programs. Providing significant improvements in diversity and difficulty, the MPII Human Pose dataset is used to train and test the ResNet50 and VGG16 models. The validation rate of the dataset is shown to evaluate the model's performance.
- 3 March 2023
- Published by Sakarya University Journal of Science in Sakarya University Journal of Science
Human Pose Estimation Using Deep Learning Models
In the last decade, there has been a lot of research into recovering 3D human poses from images. The current datasets need to address frequent exposure estimation problems adequately. However, these are shared resources to assess, inform, and contrast various models. Deep learning models are widely utilized and perform at high levels in many branches of research and engineering. So these models are used in this study with the help of the OpenCV and Keras libraries, which are open-source programs. Providing significant improvements in diversity and difficulty, the MPII Human Pose dataset is used to train and test the ResNet50 and VGG16 models. The validation rate of the dataset is shown to evaluate the model's performance.
- 1 December 2020
- Published by Emerald Publishing in Emerging Materials Research
- Vol. 9 (4) , 1266-1273
Online metallic surface defect detection using deep learning
Across a range of manufacturing contexts, automated quality control has been gaining significant attention because it offers competitive advantages such as cost reduction, high accuracy in defect detection and system stability over time. Although computer vision has been historically the most commonly applied method in this context, novel approaches such as deep learning have recently become more frequent and are used in cases where traditional methods cannot be applied. Because of the surface texture and curvature of many metallic parts, detection of defects such as scratches, cracks and dents can be challenging for traditional computer vision methods. In this study, an image acquisition system supported by a special lighting device that provides processable images from an extremely reflective cylindrical metallic surface has been developed. Multiple images obtained from a single lateral line of the surface, which is rotated at a specified speed, are combined using photometric stereo and given as input to a convolutional neural network that is employed to classify defective and non-defective samples. The results obtained from this method are close to 98.5% accurate.
- 1 December 2020
- Published by Emerald Publishing in Emerging Materials Research
- Vol. 9 (4) , 1266-1273
Online metallic surface defect detection using deep learning
Across a range of manufacturing contexts, automated quality control has been gaining significant attention because it offers competitive advantages such as cost reduction, high accuracy in defect detection and system stability over time. Although computer vision has been historically the most commonly applied method in this context, novel approaches such as deep learning have recently become more frequent and are used in cases where traditional methods cannot be applied. Because of the surface texture and curvature of many metallic parts, detection of defects such as scratches, cracks and dents can be challenging for traditional computer vision methods. In this study, an image acquisition system supported by a special lighting device that provides processable images from an extremely reflective cylindrical metallic surface has been developed. Multiple images obtained from a single lateral line of the surface, which is rotated at a specified speed, are combined using photometric stereo and given as input to a convolutional neural network that is employed to classify defective and non-defective samples. The results obtained from this method are close to 98.5% accurate.
- Open Access
- 16 July 2020
- Published by Springer Nature in Research Square
Online Metallic Surface Defect Detection Using Deep LearningOnline Metallic Surface Defect Detection Using Deep Learning
Across a range of manufacturing contexts, automated quality control has been gaining significantattention because it offers competitive advantages such as cost reduction, high accuracy in defect detection, and system stability over time. Although computer vision has historically been the most commonly applied method in this context, novel approaches such as deep learning have recently become more frequent and are used in cases where traditional methods cannot be applied. Because of the surface texture and curvature of many metallic parts, detection of defects such as scratches, cracks, and dents can be challenging for traditional computer vision methods. In this study, an image acquisition system supported by a special lighting device that provides processable images from an extremely reflective cylindrical metallic surface has been developed. Multiple images obtained from a single lateral line of the surface, which is rotated at a specified speed, are combined using photometric stereo and given as input to a convolutional neural network that is employed to classify defective and non-defective samples. The results obtained from this method are close to 98.5% accurate.
- 16 July 2020
- Published by Springer Nature in Research Square
Open AccessOnline Metallic Surface Defect Detection Using Deep LearningOnline Metallic Surface Defect Detection Using Deep Learning
Across a range of manufacturing contexts, automated quality control has been gaining significantattention because it offers competitive advantages such as cost reduction, high accuracy in defect detection, and system stability over time. Although computer vision has historically been the most commonly applied method in this context, novel approaches such as deep learning have recently become more frequent and are used in cases where traditional methods cannot be applied. Because of the surface texture and curvature of many metallic parts, detection of defects such as scratches, cracks, and dents can be challenging for traditional computer vision methods. In this study, an image acquisition system supported by a special lighting device that provides processable images from an extremely reflective cylindrical metallic surface has been developed. Multiple images obtained from a single lateral line of the surface, which is rotated at a specified speed, are combined using photometric stereo and given as input to a convolutional neural network that is employed to classify defective and non-defective samples. The results obtained from this method are close to 98.5% accurate.