Architect: extracting building related informations and changing architectural style in ar

With recent advancements in mobile computing, augmented reality (AR) applications can now achieve real-time performance. AR provides a novel way to engage with our surroundings. Platforms such as ARKit or ARCore make AR app development easy. ARchitect aims to allow users to experience different architectural styles and cultural heritages by altering the appearance of buildings in their environment. The presented solution utilizes deep learning and computer vision techniques to classify architectural styles, which is a challenging task due to the inter-class relationships between styles. The approach presented in the pa-per alters the style of a building detected from the phone's camera and retrieves information such as its current style, age, and distance from the user. To train the deep learning models, a publicly available dataset of photos of buildings with 25 different styles is used. The obtained results for architectural style classification surpass state-of-the-art methods, demonstrating the effectiveness of deep learning techniques. Additionally, the presented approach for architectural style transfor-mation is also promising and will improve as new research is conducted in image generation, monocular depth estimation, and surface reconstruction methods.