Title: Computational Methods for the Analysis of Traditional Safranbolu House Plan Typologies using Generative Adversarial Network Algorithm (GAN)

Role: Project Coordinator, Grant Recipient

Team: Taner Arsan (Researcher, KHAS), E.Füsun Alioğlu (Advisor, KHAS), Figen Özen (Advisor, Haliç Üni.), H.Fuat Alsan, Ebru Ece Tipi, Enise Nur Yılmaz, Arçın Baray Karaca

Status: Ongoing

​

Safranbolu, a UNESCO World Heritage site, boasts well-preserved Ottoman-era houses dating back to the 18th century. Despite their cultural significance, there's a lack of comprehensive research on their architectural typology. This project seeks to fill this gap by using artificial intelligence (AI) to create a database and AI-driven methods for studying these unique house designs. It aims to analyze core architectural concepts like form, symmetry, and proportion through AI applications, offering valuable insights into historical building typologies. The project's innovative approach involves training a generative adversarial network algorithm on house plan drawings, enabling both qualitative and quantitative typological analysis. By sharing its datasets and applications as open source, this project aims to inspire future interdisciplinary research in AI and architecture.