Research Focus

We focus on developing state-of-the-art AI solutions for a wide range of hot topic problems. Our projects include developing realistic behaviour models using super-agent-based algorithms and multi-agent reinforcement learning in distributed systems to train AI pilots for tactical military scenarios. These AI pilots can engage with human pilots across air, ground, and naval missions, enhancing strategic capabilities.

We are also enriching tactical environment simulators with reinforcement learning by integrating large language models to enable AI pilots to adapt to strategic commands, thus uncovering new skills and complex strategies. In healthcare, we are creating AI models supported by molecular biomarkers to accurately detect and subtype breast cancer from various imaging modalities, functioning as an early-stage second-look assistant for radiologists.

Additionally, our agricultural project leverages reinforcement learning to optimize the production of plasma-activated water, offering a natural pesticide alternative while adapting to environmental changes. Finally, we are enhancing educational experiences with a smart PTZ camera system that uses reinforcement learning to capture meaningful classroom scenes and large language models to produce comprehensive lecture notes, thereby supporting students' learning processes.

Academics

• H. Oktay Altun

Students

• Emre Fişne PhD Student

Ongoing Projects

• Developing an Ultra-Fast Artificial Intelligence Engine for Entities in Simulation Programs with Super Agent Based Algorithms and Multi-Agent Reinforcement Learning Algorithms in a Distributed System (Dağıtık Sistemde, Süper Sanal Varlık Tabanlı Algoritmalar ve Çok Etmenli Pekiştirmeli Öğrenme Algoritmaları ile Simülasyon Programlarındaki Varlıklar için Ultra Hızlı Yapay Zeka Motoru Geliştirilmesi),
  TÜBİTAK TEYDEB 1501, 2023
• Enhancing Tactical Environment Simulators with Reinforcement Learning by Enabling Virtual Entities to Adapt to Strategic Commands Using Large Language Models and Discovering New Permanent Skills to Enrich the Action Space (RL ile Akıllandırılmış Taktik Çevre Simülatörlerindeki Sanal Varlıkların LLM Modelleriyle Stratejik Komutlara Uyum Sağlayabilmesi ve Yeni Kalıcı Becerilerin Keşfedilerek Aksiyon Uzayının Zenginleştirilmesi),
  TÜBİTAK BİLGEM, 2024
• Development of an artificial intelligence model supported by molecular biomarkers for subtyping breast cancer (Meme kanserinin alt tiplendirmesine yönelik moleküler biyobelirteçler ile desteklenen yapay zeka modelinin geliştirilmesi),
  TÜBİTAK TEYDEB 1501, 2024
• Increasing Productivity and Quality in Agriculture with Reinforcement Learning Based PAW Production System (Pekiştirmeli Öğrenme Tabanlı PAW Üretim Sistemi ile Tarımda Verimlilik ve Kalitenin Artırılması),
  TÜBİTAK TEYDEB 1501, 2024
• Lecture Recording, Automatic Lecture Notes Production and Classroom Management System with Smart PTZ Camera Controller Design and Artificial Intelligence Based Algorithms (Akıllı PTZ Kamera Kontrolcüsü Tasarımı ve Yapay Zeka Temelli Algoritmalar ile Kaliteli Ders Kaydı Alma, Otomatik Ders Notu Çıkarma ve Sınıf Yönetim Sistemi)