Istanbul, Türkiye
I'm a Geomatics Engineer, recently graduated from Istanbul Technical University, focusing on remote sensing and machine learning. I'm working as a data scientist in Wipelot IoT. My toolkit includes skills in C#, .NET, SQL, JavaScript, TypeScript, React, Python, and more, giving me a strong base in software development. In addition, I also contribute as an Undergraduate Researcher where I'm merging my software development skills with my academic focus. I conduct research on GeoAI and Deep Learning at ITU Remote Sensing Laboratory, under supervision of Prof. Elif Sertel. Research Areas: - Geospatial Foundation Models - Self-Supervised Learning - Semi-Supervised Learning - Quantum Machine Learning https://github.com/canmike
• Conducting research on GeoAI at ITU Remote Sensing Laboratory, under supervision of Prof. Elif Sertel. • Research Areas: Geospatial Foundation Models, Self-Supervised Learning, Semi-Supervised Learning, Explainable AI. • Published a research paper titled “Automatic Road Extraction from Historical Maps Using Transformer-Based SegFormers” in the ISPRS International Journal of Geo-Information. • Poster presentation “Benchmarking Foundation Models for Land Cover Segmentation Using DeepGlobe Land Cover Dataset” at ESA-NASA International Workshop on AI Foundation Model for EO 2025. • Oral Presentation: “Zero-Shot Land Cover Segmentation with GroundingDINO-Segment Anything Model on WorldView-3 and DeepGlobe Data” at MIGARS 2025. • Oral presentation: “Evaluating Generative SAR-to-RGB Completion in Earth Observation with Performance, Uncertainty, and Explainability” at ML4Earth AI4EO 2025 Workshop. • Scientific and Technological Research Council of Turkey (TÜBİTAK)-funded project titled “Automatic Detection of Craters on the Surface of Mars Using Semi-Supervised Deep Learning Methods”
• Modified the YOLOv8 architecture to enable object distance estimation. • Deployed a lightweight YOLOv8n model on a Raspberry Pi AI camera for real-time edge inference. • Designed a custom time-based clustering algorithm, prioritizing locations with prolonged dwell times for more context-aware grouping.
• Developed and maintained 3D interface and 2D map on web for displaying real-time IoT device locations. • Implemented using TypeScript, React, Three.js, Redux Toolkit and SignalR.
• Led a team of 8 developers, working on Swarm UAV technologies. • Designed the software architecture of Swarm UAV project using Python and ROS. • Developed and trained a language model architecture using LSTM Encoder-Decoder for Sequence Labeling of swarm formations. • Developed a Vision-Language Model for Swarm UAVs by fine-tuning CLIP and GPT-2, guiding the team through this complex project to enable drones equipped with cameras to understand and respond to written queries and commands.
Working on a project funded by the European Research Council, GeoAI_LULC_Seg: A GeoAI-based Land Use Land Cover Segmentation Process to Analyse and Predict Rural Depopulation, Agricultural Land Abandonment, and Deforestation in Bulgaria and Turkey, 1940-2040. The GeoAI_LULC_Seg project aims to employ geospatial AI to accurately map LULC from 1940 onward in a border region between Bulgaria and Turkey. https://cordis.europa.eu/project/id/101100837