Oya Kilci

Dr. Oya Kilci

orcid

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
25 January 2025 in Food Analytical Methods
Verified
Journal Article
Open access
PDF file
OKOya Kilci
Oya Kilci
Murat KOKLUMurat KOKLU
Murat KOKLU
It is of great importance for food safety and consumer satisfaction that industrial food products are durable, hygienic, and flawless. Robust products protect the physical integrity of the product by preventing damage that may occur during the production and transportation processes, which meets the expectations of the consumer. Hygienic production conditions prevent foodborne diseases by minimizing the risk of microbial contamination and protect consumer health. Perfect products strengthen the brand image with their aesthetic and satisfactory features and increase consumer loyalty. In the study conducted in this context, the classification of defect and no defect conditions of biscuits in the food industry was examined using YOLOv8 models. A summary dataset consisting of 4990 biscuit images was created and the biscuits were initially divided into two categories: defect and no defect. Later, defect biscuits were classified into three subcategories: not complete, overcooked, and texture defect. As a result of experiments with YOLOv8 models, binary classification (defect, no defect), the highest accuracy rate was achieved in the YOLOv8-m, YOLOv8-l, and YOLOv8-x models with 96.78%, while the highest accuracy rate in the triple classification (not complete, overcooked, and texture defect) performance was achieved in the YOLOv8-m model with 96.99%.

26 September 2025 in Research in Agricultural Sciences
Verified
Journal Article
Open access
OKOya Kilci
Oya Kilci
Murat KOKLUMurat KOKLU
Murat KOKLU
This study presents a classification approach for guava fruit diseases using both deep learning and machine learning models. InceptionV3 was employed to extract image features, which were subsequently classified using models such as artificial neural networks support vector machines, k nearest neighbors, random forest, and decision tree. The performance of the models was evaluated in terms of accuracy, F1 score, precision, and recall. Experimental results demonstrate that SVM and ANN achieved the highest performance, with SVM reaching 0.9974 across all metrics and ANN achieving 0.9958. The kNN model also performed well with an accuracy of 0.9924, while random forest and decision tree obtained lower accuracies of 0.9612 and 0.9209, respectively. Confusion matrix analysis further confirmed the superiority of SVM and ANN, with minimal misclassifications across anthracnose, fruit fly, and healthy guava categories. These findings highlight the effectiveness of deep learning-based feature extraction combined with SVM and ANN classifiers for reliable and accurate detection of guava fruit diseases.

17 July 2025 in Journal of Food Science
Verified
Journal Article
Closed access
OKOya Kilci
Oya Kilci
YEYusuf Eryeşil
Yusuf Eryeşil
Murat KOKLUMurat KOKLU
Murat KOKLU

31 December 2019 in International Journal of Applied Mathematics Electronics and Computers
Verified
Journal Article
Open access
PDF file
OKOya Kilci
Oya Kilci
Murat KOKLUMurat KOKLU
Murat KOKLU
The need for energy is increasing by the day. Renewable energy sources already have a significant place in electricity production both around the world and also in Turkey and the share of solar energy going higher within the renewable energy sources. Solar energy is extremely easy to use, it is preferable too because it is environmentally friendly-clean energy. Solar energy plant is the energy source of the future, because the production costs of solar plants are lower than other energy sources and the economic difficulties are overcome. To get optimal production, the solar energy plant must make maximum use of the sun. In the designed system, solar panels are provided with both fixed and seasonal variations. The first variations of solar panel systems is are placed at constant 25° angle and the position angle of second variation is shifted from 10° to 30°in summer and winter accordingly. In this study, the performance of 1148 kWp PV plants which is established in Altınekin district of Konya and planned to be connected to the grid was evaluated by PVSYST analysis program. Power generation of the two different systems, specific data and performance value are calculated. Result of the study shows that the seasonal system produced 32.2 MWh more per year than the fixed system. Global irradiation in the seasonal system is 26.4 kWh/m², and the specific yield is concluded to be more than 28 kWh / kWc / year.

31 December 2025 in Intelligent Methods in Engineering Sciences
Verified
Journal Article
Open access
PDF file
IAIbrahim Akbas
Ibrahim Akbas
OKOya Kilci
Oya Kilci
Murat KOKLUMurat KOKLU
Murat KOKLU
This study aims to develop a convolutional neural network (CNN)-based classification framework that can distinguish between synthetic and real animal images generated by the Stable Diffusion Turbo model. Additionally, this study will evaluate the performance of different network architectures for this task. The study employed a balanced dataset of 31,995 images, including 16,000 real images and 15,995 synthetic images generated by Stable Diffusion Turbo. The dataset includes eight animal categories: dogs, cats, cows, rabbits, horses, sheep, chickens, and elephants. All images were resized to 224 by 224 pixels, and standard preprocessing techniques were applied. During the classification stage, five pretrained convolutional neural network architectures were retrained using transfer learning, including MobileNetV2, DenseNet121, DenseNet169, DenseNet201, and NASNetMobile. Model performance was evaluated using accuracy, precision, recall, the F1 score, the area under the curve of the receiver operating characteristic, confusion matrices, and training time. The experimental results demonstrate that MobileNetV2 and DenseNet201 achieved the highest classification performance, with respective accuracy rates of 99.58% and 99.56%, and perfect area under the curve values. All DenseNet variants exhibited complete sensitivity in detecting synthetic images, whereas NASNetMobile showed substantially lower performance compared to the other models. These results suggest that synthetic images produced by diffusion-based generative models can be reliably identified when appropriately designed CNN architectures and balanced datasets are used. This provides a significant methodological contribution to the discrimination of synthetic versus real images, the detection of fake content, and the verification of visual authenticity.

12 December 2025 in Proceedings of International Conference on Intelligent Systems and New Applications
Verified
Journal Article
Open access
PDF file
MSMucahid Mustafa Saritas
Mucahid Mustafa Saritas
OKOya Kilci
Oya Kilci
Murat KOKLUMurat KOKLU
Murat KOKLU
In this study, the gear fault classification problem, which is of critical importance in industrial mechanical systems, was investigated within the scope of five deep learning models including ResNet18, ResNet34, ResNet50, DenseNet121 and EfficientNet-B0 architectures widely used in the literature. Models were trained on the multi-class gear fault image dataset and their accuracy performances were compared with their numerical values. According to the results, ResNet18 achieved the highest accuracy value with 0.9615, while EfficientNet-B0 showed a similarly strong performance with 0.9594. ResNet34 ranked third with an accuracy value of 0.9541, demonstrating that lightweight ResNet architectures offer high generalization ability in gear fault detection. On the other hand, deeper architectures, ResNet50 with 0.7511 accuracy and DenseNet121 with 0.7500 accuracy, did not provide a significant increase in accuracy despite increasing structural complexity and showed limited performance against the characteristics of the data set. These findings reveal that representation efficiency rather than model depth is the determining factor in gear fault classification problems, and that ResNet18 and EfficientNet-B0 architectures are the most suitable options for real-time fault detection systems.