Skip to content
DSML.KZ
Publications

Talk · International Conference on Computer Systems and Technologies · June 14, 2024

Developing an Artificial Intelligence Tool for Image Generation Using a Unique Dataset with Image-to-Image Functionality

Akbiike Toktosunova, Alymbek Ergeshov, Gulzada Esenalieva, Andrei Ermakov, Ruslan Isaev

Ruslan Isaev
Open via DOI

Abstract

This article describes the application of a unique artificial intelligence model, trained on the basis of a special dataset, for drawing images. Artificial intelligence is very important in the modern world because it is a key driver of progress, offering innovative solutions to improve processes in various fields. A feature of this work is the use of the Stable Diffusion model and the creation of the LoRa model to create a specific object. The use of a special dataset, which is a classified set of images, to train models based on a stable propagation mechanism. The application of the image-to-image method, which consists in completely generating an image by having the model reproduce all the objects in the image, trying to preserve the maximum amount of detail in the image. As a result, a tool has been developed that can generate images of characters of a certain style.

Generative Adversarial Networks and Image SynthesisAdvanced Neural Network ApplicationsDigital Media Forensic Detectionimage generationStable DiffusionLoRa

More by Ruslan Isaev