Featured technical projects I've built
Computes Cardiothoracic Ratio (CTR) and a cardiomegaly score from chest X-rays. Uses XRV DenseNet121 for the CM score and Ultralytics YOLOv11 for lung/heart width-based ratio calculation, with CLAHE, NMS, and GPU acceleration. Results are saved to JSON.
🏆 Teknofest AI in Healthcare – 4th in Turkey
Detects polyps at pixel level in colonoscopy images. Uses YOLOv8-seg for inference and a DeepLabV3 teacher → student (knowledge distillation) setup for training. Outputs (masks/boxes + metrics) are saved to JSON.
🏆 Teknofest AI in Healthcare – 4th in Turkey
Made an electric vehicle autonomous on ROS2 Foxy and Docker. Built sensor fusion with LiDAR, IMU, GNSS; an end-to-end stack with Perception, SLAM, Localization, Planning, Control, path tracking, and obstacle avoidance modules.
🥉 Teknofest Robotaxi – 3rd in Turkey (Honorable Mention)
A deep learning model that detects spinal landmarks (key points) in X-ray images and automatically computes the geometric Cobb angle, built using the AASCE MICCAI 2019 dataset and ResNet50 architecture.
🔬 Independent Research Project (2026)