Fatih Furkan ARSLAN
@bozok.edu.tr
Research Assistant, Department of Electrical and Electronics Engineering, Faculty of Engineering and Architecture
Yozgat Bozok University
@bozok.edu.tr
Research Assistant, Department of Electrical and Electronics Engineering, Faculty of Engineering and Architecture
Yozgat Bozok University
This project aims to develop a high-performance computer vision system for automatic detection and classification of microscopic seed images using deep learning. The study will integrate YOLO-based object detection models with classification architectures such as ResNet, EfficientNet, and MobileNet. Additionally, segmentation techniques like Segment Anything Model (SAM) will be explored to improve accuracy. The project focuses on handling class imbalance, optimizing model performance, and developing a robust pipeline suitable for real-world agricultural and biological applications.
This project focuses on developing a real-time motion classification system using IMU sensor data and machine learning techniques. The system will analyze time-series data from accelerometer and gyroscope sensors to classify motion patterns. The project includes feature engineering, deep learning models such as 1D CNN, and deployment on embedded platforms like ESP32. The goal is to build an efficient, low-latency system suitable for real-time applications.
This project aims to design and evaluate hybrid and ensemble deep learning models to improve classification performance on complex image datasets. Multiple architectures such as ResNet, EfficientNet, and MobileNet will be combined using ensemble techniques and cross-validation strategies. The project will focus on improving macro-F1 scores, reducing overfitting, and ensuring robust generalization across datasets.
Scholar Citations
Scholar h-index
