Modern Neural Network Technologies Text-to-Image

Abstract

This paper discusses state-of-the-art graphical text-to-image neural networks and methods for text-to-image conversion, analyzing the results achieved and samples created to date for text-to-image conversion tasks. Ways of applying neural network approaches to text-to-image transformation for environmental monitoring, infrastructure and medical data analysis tasks are proposed. In this paper the results of neural network generation and its correlation with the user input linguistic constructions of text queries are reviewed, and the typical flaws and artifacts typical of the neural network generated images are identified and classified. The rapid development of neural network technologies in this field could have a significant impact on society, the professional market and the media, which makes the task of studying neural network images and identifying them among other graphic content particularly relevant.