CREATING ARTIFICIAL IMAGES AND VIDEO USING GENERATIVE NEURON NETWORKS (GAN, DIFFUSION MODELS)

Abstract

The article analyzes the capabilities of generative neural networks in the process of creating images and videos, and especially considers the most important technical issue encountered in video generation - the problem of time consistency disruption. The proposed solution creates the possibility of wide application for the automation of video generation in the creative industries, film, design, and the production of artificial multimedia.