Crosspower

In the context of the modern information war, the development of effective methods to counter disinformation and hostile propaganda has become critically important. The amount of false information and the speed of its dissemination necessitated the implementation of automated systems involving artificial intelligence to optimise the processes of creating visual counter-propaganda content. This research aimed to develop a methodology for the integration of artificial intelligence technologies into the processes of creating effective visual tools for countering disinformation, taking into account the principles of graphic design and the psychology of visual information perception. The research was based on a comprehensive approach that combined theoretical analysis of scientific literature, comparative analysis of the neural networks Midjourney, Stable Diffusion, and DALL-E, semiotic analysis of visual materials, as well as experimental implementation of the developed system for the automated creation of counter-propaganda visual content. A comprehensive approach has been developed for the creation of visual tools to counter disinformation, which combines the capabilities of automated information collection systems, artificial intelligence algorithms for generating graphic content, and the principles of effective graphic design. It has been found that the use of artificial intelligence in graphic design has optimised up to 20% of routine tasks in the creation of visual content, allowing designers to focus on the strategic and creative aspects of development. The developed recommendations for the use of artificial intelligence in graphic design may be implemented by state institutions, media, and public organisations to respond promptly to information threats

This article is part of a larger study of design as a cultural phenomenon. In this part of the study, the author examines the process that is currently taking place in the semiotic structure of design, associated with the active introduction of neural networks into the creation of visual content. Artificial intelligence products stylistically take the design away from the fourth-order simulacrum (from the flat design style) and return the design to using the third-order simulacrum as the main iconic form.The object of the research is the transformation of the semiotic design system. The subject of the study is a return to the third–order simulacrum in modern design. The purpose of the study is to show and try to explain how in modern design there is a return to the use of a third-order simulacrum. The research method is a semiotic analysis of modern design based on the methodology of R. Barth's semiotic analysis. The study is also based on J. Baudrillard's theory of semiosis in hyperreality and three orders of simulacra. The author sees the philosophical justification of the art of neural networks in the concept of flat ontologies. The study of the semiotic structure of design allows us to see that the logic of design development carries this phenomenon through successive stages of semiosis associated with a decrease in meaning and a diminution of being. In its development, the design consistently uses first index signs, a second-order simulacrum, then a third-order simulacrum as the main sign form. Nowadays, the main iconic form in design has become a fourth-order simulacrum. Next, design had to either end as a profession and phenomenon, or move into a new cultural paradigm that was not related to simulation. However, unexpected transformations have begun to occur in design, due to the active involvement of non–human agents - neural networks – in the creation of visual content. Neural network products are a typical example of a third-order simulacrum. Thanks to the use of neural networks, modern design finally acquires the vector of transhumanism and closes in simulation.

To date, the use of applications for the generation of illustrative material with the help of artificial intelligence (AI) is one of the most progressive for creating content in the field of visual occupations (designers, architects, artists), for marketers, students and ordinary people. The main reason for using neural networks is to save time and create inspiring examples in any field of human activity. Currently, there are more than 20 independent programs that generate visual content, and many companies such as Adobe and Canva use neural network tools. The use of artificial intelligence is irreversible and requires practice and some experience in its work with the creation of visual content. The development and updating of the main programs in this direction, such as Midjourney and Leonardo, is constant and needs to be studied. The photo-realism and image detailing of the latest versions of Midjourney allows you to create visual content that can be used to generate new ideas and use it for creation quality visual content for advertising. The article compares the quality parameters of the visual content created on the basis of artificial intelligence by the Midjourney program using its different versions. The updates from the initial version to the latest V6.0 were phased in less than a year and a half. Certain deviations in the images are considered, which leads to the impossibility of their further use. Analyzed the parameters of writing explanations (promts), which will affect the final quality of the generated image, the possibility of writing complex promts with one and several images, as well as advanced promts. The possible use of neural networks that work with text for targeted writing of appropriate prompts for better automation of the generation process is considered. The article considers the possible use of the Midjourney program for various spheres of human activity that use various images, and also raises the question of the role of a person as the main creator and generator of creative ideas.

There is a widening gap between the creation of visual content and its analysis and interpretation by machine, an increasingly essential require-ment for correct indexing and filtering. In the case of the WWW, for instance, although there are efficient methods to process the encoded (e.g. ASCII) text, there are no such methods for the (significant) visual content. This paper focuses on the methods developed by the authors to address the problem of extracting the characters from WWW images containing text.

The study explored the educational potential of the application of student-generated digital visual content for learning English as a second language (ESL) by undergraduate students enrolled in the course Foreign Language which is actually Introduction to Legal English. This study used a mixed-methods approach. The researchers designed a quasi-experimental design to examine whether the students’ creation of visual content, supported by structured use of artificial intelligence (AI), could improve second language learning outcomes, increase motivation, and promote critical engagement with digital tools. The experimental group was tasked with creating personalized visual learning materials. The applied approach was structured in several steps, from creating simple forms including infographics and comparative charts to poster presentations and digital video passion projects. The algorithm for collaboration with AI and the work with specific features of AI-generated materials was applied aimed at making a student a critical consumer of this content and mitigating potential drawbacks of using AI. To assess the learning outcomes after the intervention, the post-test was administered, which revealed that the studied instructional design had a positive impact on language development across all aspects checked. The questionnaire, which included both open-ended and closed-ended questions, investigated students’ perceptions of the applied methodology and faced challenges. The findings showed that students perceived integrating visual creation and structured AI-supported activities into English language learning as beneficial for language skills development, boosting motivation and interest, and the advancement of digital literacy.

The study explored the educational potential of the application of student-generated digital visual content for learning English as a second language (ESL) by undergraduate students enrolled in the course Foreign Language which is actually Introduction to Legal English. This study used a mixed-methods approach. The researchers designed a quasi-experimental design to examine whether the students’ creation of visual content, supported by structured use of artificial intelligence (AI), could improve second language learning outcomes, increase motivation, and promote critical engagement with digital tools. The experimental group was tasked with creating personalized visual learning materials. The applied approach was structured in several steps, from creating simple forms including infographics and comparative charts to poster presentations and digital video passion projects. The algorithm for collaboration with AI and the work with specific features of AI-generated materials was applied aimed at making a student a critical consumer of this content and mitigating potential drawbacks of using AI. To assess the learning outcomes after the intervention, the post-test was administered, which revealed that the studied instructional design had a positive impact on language development across all aspects checked. The questionnaire, which included both open-ended and closed-ended questions, investigated students’ perceptions of the applied methodology and faced challenges. The findings showed that students perceived integrating visual creation and structured AI-supported activities into English language learning as beneficial for language skills development, boosting motivation and interest, and the advancement of digital literacy.

This study investigates how visual content influences marketing effectiveness on LinkedIn, focusing on its role in enhancing engagement, improving brand perception, and amplifying message clarity. As LinkedIn evolves from a professional networking site to a comprehensive platform for B2B communication and digital branding, organizations are increasingly leveraging it to share updates, build authority, and foster professional relationships. Visual content—such as infographics, short videos, graphics, carousels, and animations—has emerged as a key tool in capturing user attention and improving the delivery of complex information. The research combines both primary and secondary data sources. A survey was conducted among professionals from sectors including IT, marketing, education, and human resources, aiming to understand how visual elements affect content preference, engagement behaviour, and memory retention. Findings suggest that visual content significantly outperforms text-based posts in driving interactions such as likes, comments, shares, and click-through rates. Infographics are particularly effective for simplifying data, while short-form videos are favored for their ability to communicate brand value quickly and persuasively. Supporting secondary literature and LinkedIn’s algorithmic patterns confirm that posts enriched with visual media tend to receive broader organic reach. Studies also highlight that visual storytelling increases brand recall and strengthens trust and credibility. Furthermore, neuroscience supports the idea that visuals are processed faster than text, making them a powerful asset for marketers aiming to deliver impactful messages in a short time. The paper also explores strategic applications of visual content on LinkedIn, recommending the use of consistent branding, high-quality design, and goal-oriented visuals tailored to specific marketing objectives—such as employer branding, lead generation, or thought leadership. It encourages investment in design resources and training to maximize the effectiveness of visual marketing. However, the study recognizes certain limitations, including a geographically narrow sample (primarily Indian professionals), a relatively small data pool, and reliance on self-reported user behaviour. The absence of experimental methods like A/B testing also limits the ability to establish direct causation. In conclusion, the research calls for future studies on the integration of emerging technologies like AI in visual content creation, the development of platform-specific visual strategies, and the importance of accessibility and inclusivity in design. As digital engagement continues to shift towards visual-first experiences, LinkedIn marketers must adapt to stay competitive and relevant in the professional content ecosystem. Keywords: Visual Content, LinkedIn Marketing, User Engagement, Brand Visibility, Content Strategy, Infographics, Short-form Videos, B2B Communication, Professional Networking, Social Media Marketing, Brand Recall, Visual Storytelling, Digital Branding, Click-through Rate, Organic Reach, Marketing Analytics, Content Performance, Thought Leadership, Employer Branding, Visual Design

Abstract: The “AI-Powered Image Generation System using AI” is designed to transform textual descriptions into visually compelling images using advanced artificial intelligence techniques. By leveraging AIstate-of-the-art models, such as DALL•E, the system interprets user provided prompts and generates high-quality, realistic, and creative images in real-time. This approach demonstrates how AI can bridge human imagination and machine intelligence, making creative content generation faster and more accessible. The system architecture consists of a user-friendly interface for inputting prompts, a backend module that communicates with the AI API, and an image rendering component that delivers generated images to the user. The platform allows both technical and non-technical users to create visuals efficiently, supporting applications in digital art, marketing, education, and entertainment. By integrating AI into the creative workflow, the project highlights the potential of prompt-based image generation in modern content creation. In addition, the project addresses challenges such as handling ambiguous or inappropriate prompts, ensuring image quality and resolution, and promoting ethical use of AI-generated content. Future developments may include multi-modal input, style customization, and integration with other creative tools. Overall, this project illustrates the transformative capabilities of AI in visual content creation and provides a scalable framework for innovative applications across various domains..

Visual content creation has spurred a soaring interest given its applications in mobile photography and AR / VR. Style transfer and single-image 3D photography as two representative tasks have so far evolved independently. In this paper, we make a connection between the two, and address the challenging task of 3D photo stylization - generating stylized novel views from a single image given an arbitrary style. Our key intuition is that style transfer and view synthesis have to be jointly modeled for this task. To this end, we propose a deep model that learns geometry-aware content features for stylization from a point cloud representation of the scene, resulting in high-quality stylized images that are consistent across views. Further, we introduce a novel training protocol to enable the learning using only 2D images. We demonstrate the superiority of our method via extensive qualitative and quantitative studies, and showcase key applications of our method in light of the growing demand for 3D content creation from 2D image assets.

Image generation allows the creation of visual content in a convenient manner. It is critical for enhancing digital experiences, from video games to virtual reality, enabling more engaging and immersive experiences. In current technologies, Generative Adversarial Networks (GANs) have achieved significant success but face challenges like training instability and mode collapse. By utilizing the Wasserstein distance, Wasserstein GAN (WGAN) enhances conventional GANs; however, its weight clipping method may not be ideal. In this study, WGAN with gradient penalty (WGAN-GP) and WGAN with weight clipping (WGAN-CP) are compared, which aims to enhance stability by better enforcing the Lipschitz constraint. For comparison, these approaches are validated using Fashion Modified National Institute of Standards and Technology (MNIST) and Canadian Institute for Advanced Research (CIFAR)-10 datasets. Experimental results show WGAN-GP produces higher quality images and more stable training than WGAN-CP. However, WGAN-GP also requires longer training times and computational burden. The findings highlight a trade-off between training efficiency and output quality, guiding the choice of technique based on specific application needs.

Shape Correspondence and Functional Maps

Notions of similarity and correspondence between geometric shapes and images are central to many tasks in geometry processing, computer vision, and computer graphics. The goal of this course is to familiarize the audience with a set of recent techniques that greatly facilitate the computation of mappings or correspondences between geometric datasets, such as 3D shapes or 2D images by formulating them as mappings between functions rather than points or triangles. Methods based on the functional map framework have recently led to state-of-the-art results in problems as diverse as non-rigid shape matching, image co-segmentation and even some aspects of tangent vector field design. One challenge in adopting these methods in practice, however, is that their exposition often assumes a significant amount of background in geometry processing, spectral methods and functional analysis, which can make it difficult to gain an intuition about their performance or about their applicability to real-life problems. In this course, we try to provide all the tools necessary to appreciate and use these techniques, while assuming very little background knowledge. We also give a unifying treatment of these techniques, which may be difficult to extract from the individual publications and, at the same time, hint at the generality of this point of view, which can help tackle many problems in the analysis and creation of visual content. This course is structured as a half day course. We will assume that the participants have knowledge of basic linear algebra and some knowledge of differential geometry, to the extent of being familiar with the concepts of a manifold and a tangent vector space. We will discuss in detail the functional approach to finding correspondences between non-rigid shapes, the design and analysis of tangent vector fields on surfaces, consistent map estimation in networks of shapes and applications to shape and image segmentation, shape variability analysis, and other areas.

The development of digital technology has changed the way students interact with learning media, including in the creation of visual content. One of the most widely used applications among students is Canva Mobile, which offers convenience in designing various academic needs practically through mobile devices. This research aims to explore the experiences of second-semester students of the Communication Science Study Program at the University of Boyolali in using the Canva Mobile interface as a supporting medium for completing academic assignments. This study uses a descriptive qualitative approach with data collection techniques through in-depth interviews and observation. Informants were selected using purposive sampling, followed by snowball sampling. The collected data were analyzed through data reduction, data presentation, and conclusion drawing. The results show that the majority of students feel comfortable and find it easy to use Canva Mobile. Its user-friendly interface, diverse features, and flexible access are the main reasons for its popularity. Canva Mobile has been proven to improve the quality of students' visual assignments and support creativity development in the field of visual communication. However, there are some obstacles, such as the limited features in the mobile version compared to the desktop version, as well as restricted access to premium elements. This study concludes that Canva Mobile is an effective, adaptive, and relevant learning medium that aligns with the needs of today’s digital-native students. Students' positive experiences in using this application serve as evidence that mobile technology can be optimally integrated into higher education.

Role of the development of digital technologies in the creation of visual content

In the context of education globalization, the importance of a foreign language communicative competence for a specialist and potential employee is indisputable. For effective work in the future, the current student should master the skills of analyzing, working out and creation of visual content, hence one of the preconditions for a specialist’s success (first of all, in technical specialty) is the ability to quickly process and present large information volumes in the most accessible and compressed form. The fundamentals of productive skills mentioned above are the skills of visualized content analysis and representing in different formats trained in university courses, especially in a foreign language course. This article is devoted to the study of the characteristics of modern visualization means and their potential in the educational process of the foreign language students of non-linguistic specialties of the bachelor program. The reasons for motivation increase necessity in class activities are presented; the importance of training students for self-directed production of independent analytical message in their future professional activity is emphasized. The advantages of working out visualized foreign data over the data in text format based on a student survey were determined. The students’ perception of the visualized content which is integrated into English textbooks and typical of the world’s leading educational publishing houses, in the view of necessity, self-descriptiveness and aesthetics was analyzed. In the article, a strategy for linguistic aspect training for graphic information processing by students and for creation of their own oral or written discourse in accordance with the certain learning objectives or speech intentions is drawn up. The preconditions for successful visualization tools’ integration into class activities under the teacher guidance are presented.