Noise reduced SAR ship database
8th International Artificial Intelligence and Data Processing Symposium
Introduces a noise-reduced dataset for SAR ship detection and benchmarks annotated vessel imagery for research use.
Research Assistant · PhD, İnönü University
Taha Burak ÖzdemirDeep learning for SAR image analysis & remote sensing.
I'm a PhD candidate at İnönü University researching deep learning for ship detection in Sentinel-1 SAR imagery — with secondary work on autonomous multi-agent systems and reproducible ML engineering.
Profile
About
I'm a Research Assistant in the Department of Computer Engineering at İnönü University, where I am pursuing my PhD in Computer Engineering. My doctoral research investigates deep learning for ship detection in Sentinel-1 SAR imagery, with a focus on speckle-noise robustness and class imbalance in Earth-observation datasets.
Outside the lab, I work across the full software stack — from production React/Next.js applications to data pipelines in Python — which I see as essential for turning research artifacts into systems that actually run.
I'm interested in the intersection of remote sensing, autonomous multi-agent systems, and the engineering practices that make ML research reproducible.
Current research
SAR ship detection in Sentinel-1 imagery
Investigating deep learning approaches for maritime object detection in Synthetic Aperture Radar data, with a focus on speckle-noise robustness and class imbalance.
DomainEarth observation
MethodDeep learning
ProgramPhD '24 – '27
Open toResearch collaboration
Research output
Publications
A growing list of peer-reviewed publications. Full list and citation graph on Google Scholar.
SAR ship detection based on differential image analysis and machine learning approach
2024 publication
Presents a differential image analysis pipeline with machine learning classifiers for improved SAR ship detection accuracy.
Ship classification based on co-occurrence matrix and support vector machines
Electrica 24 (3), 812-817
Classifies maritime targets using texture features from co-occurrence matrices and support vector machine models.
SAR ship detection using image histograms and machine learning approach
Türk Doğa ve Fen Dergisi 13 (3), 171-175
Uses image histogram descriptors with traditional machine learning to detect ships in SAR imagery.
Selected work
Projects
A selected set of projects spanning research, intelligent systems, and applied software engineering.
SAR ship detection
Deep-learning pipeline for ship detection in Sentinel-1 SAR imagery, including dataset construction with GDAL and ESA SNAP toolkits, speckle-noise pre-processing, and detection benchmarks.
Python · PyTorch · GDAL · SAR · ESA SNAP · Computer Vision
Autonomous drone coordination
Distributed drone coordination, task allocation, and autonomous mission planning using AI-driven decision mechanisms across heterogeneous agents.
Python · Multi-Agent Systems · ROS · Reinforcement Learning
Code under preparation — paper in review
Side projects
Production-leaning side work that complements the research above.
Movie recommender
A personalized movie recommender with content-based filtering and a responsive React/Next.js front-end, backed by an Express + MongoDB Atlas service.
React · Next.js · Node.js · Express · MongoDB Atlas · Axios
Academic journey
Education
Software engineering, intelligent systems, and research-driven work in artificial intelligence and computer vision.
2024 – Present
PhD in Computer Engineering
İnönü University · MalatyaDoctoral research on artificial intelligence, computer vision, SAR image processing, and multi-agent systems.
2022 – 2024
MSc in Software Engineering
İnönü UniversityThesis-track research on machine learning and software systems with an academic publication focus.
2018 – 2022
BSc in Software Engineering
Bahçeşehir University · İstanbulFoundations in software engineering, algorithms, distributed systems, and applied ML.
Tools
Tools & Stack
What I reach for across research, intelligent systems, and modern web/data work.
- Research
- Python · PyTorch · NumPy · OpenCV · GDAL · ESA SNAP
- Engineering
- TypeScript · React · Next.js · Node.js · PostgreSQL
- Infrastructure
- Docker · Git · Linux · GitHub Actions
- Domain
- SAR / Sentinel-1 · Remote Sensing · Computer Vision · Multi-Agent Systems
Contact
Open to research collaborations.
Academic collaborations, research discussions, software projects, or professional opportunities — feel free to get in touch.
Emailtahaburak@tahaburakozdemir.comBest for academic and professional contactGoogle Scholarscholar.google.com/citations?user=tWFlGgYAAAAJPublications & citationsGitHubgithub.com/tburakozdemirProjects, code, and technical workLinkedInlinkedin.com/in/taha-burak-ozdemirAcademic and professional network
Open to research collaborations & projects
Send a message