USE OF ARTIFICIAL INTELLIGENCE IN AUTOMATED INTERIOR DESIGN TASKS

The paper analyses current trends in interior design. It is shown that the most promising way to take into account the personalized requirements of customers and the variety of interior objects is the development of automated systems. This will allow to increase work efficiency by reducing the burden on designers, reducing task completion time, and allowing to make changes to projects more quickly. For the operation of such automated system, it is necessary to develop appropriate computer-aided interior design tools and human-machine interface software for inputting data and saving results. Based on the 3D-FRONT dataset, the ATISS neural network was configured for the task of automated design of kitchens and living rooms, taking into account the style of the rooms and the number of things. A web application for interaction with the user has been developed, which consists of a server part on FastAPI and a front-end part on Gradio. The user can upload photos of rooms to generate interiors and get a 3D model of the room as a result.