Pozitiva a negativa generování obrázků pomocí AI

This bachelor thesis deals with the issue of image generation using artificial intelligence. The opening section describes the basic principles of artificial intelligence and provides an overview of key generative models focused on the creation of visual content. The thesis focuses on the most widely used tools, DALL-E, Midjourney, Stable Diffusion, and Adobe Firefly, and compares them in terms of accessibility, and user control options. The theoretical part also covers the issue of prompt engineering and provides a detailed analysis of the pros and cons of using generative AI for image creation. The practical part of the thesis is based on a questionnaire survey focused on the respondents' ability to recognize the difference between images created by humans and images generated by artificial intelligence. Respondents answered binary questions as well as multiple-choice questions. The results were analyzed using statistical methods, including correlation, t-tests, and one-way ANOVA. The analysis shows differences between task types, the influence of experience with AI and visual creation on correct answers, and the relationship between subjectively perceived difficulty and actual accuracy. The work thus provides a comprehensive view of the possibilities and limitations of visual generative AI and contributes to a better understanding of how people evaluate and interpret visual content created by modern generative models.