Projects
Current and completed work in visual UAV navigation, terrain analysis, artificial intelligence and research software.
Active
The project examines morphological change, surface retreat and rockfall hazards affecting Cappadocia’s fairy chimneys and tuff formations, with the goal of producing an inventory and a risk-priority map.
Active
This project adapts a ResNet50 regression model to estimate UAV altitude from nadir imagery collected under varied terrain, weather and illumination conditions.
Active
The project investigates automatic archaeological site detection from digital elevation models and high-resolution orthophotos using deep learning and terrain-aware image analysis.
Active
This study examines whether patterns in student achievement data can support evidence-based identification of programme orientation and improve academic guidance.
Completed
The project developed and integrated a web-based student assistant that uses a tailored OpenAI model to provide accessible, context-aware academic support.
Completed
The doctoral research evaluates terrain-based visual localization for UAVs by learning compact map representations with autoencoders and matching nadir imagery when GNSS is unavailable.
Completed
The project coordinated the development of administrative automation software to streamline course, participant and operational records for Tepebaşı Public Education Centre.