Convolutional Neural Network in the Classification of Architectural Styles of Buildings
Eldar Dautov, N.A. Astafeva · 2021
The article focuses on the implementation of the project for the classification of architectural styles of buildings. We use modern deep learning methods using convolutional neural networks for realization of this project. Image classification using CNN has many implementations. However, the recognition of architectural styles is not fully researched. This article presents a neural network capable of determining from a photograph of a building its belonging to a certain style of architecture. We train a deep CNN classifier model to classify buildings belonging to 15 different architectural styles. We use a data set that we have collected from open sources of the global Internet to do this.