Recent developments in the field of Artificial Intelligence (AI) and more specifically Deep Learning (DL), are gradually contributing to the study and classification of the typology of Cultural Heritage (CH) buildings. A method based on the latest developments in deep learning is presented to classify modern Athenian architecture (from 1830) using Artificial Neural Network of type YOLO11. YOLO11 is trained to identify the typology of various Athenian buildings from their facade, identifying distinctive morphological characteristics and structural elements of each period. The method is tested and validated using a large dataset of more than 3500 images of Athenian buildings, that includes multiple instances of buildings categorized into five distinctive classes, namely, neoclassical, neoclassical-eclectic, interwar, interwar -eclectic and postwar buildings. This research may contribute to the digital management of the Athenian architectural heritage, creating the core methodology for large-scale categorization from real-time street images, videos or from data collected by UAVs. Moreover, the classification of buildings is a key source of knowledge in urban and spatial city planning, which decisively determines not only the protection zones and land uses but for infrastructure management, financial planning, the emergency planning, as well as actions for climate change adaptation and heritage preservation.