In this thesis, a detailed analysis of the approach to applying generative artificial intelligence (genAI) for product concept visualization is presented. The primary goal was to explore the possibilities and limitations of various genAI tools that enable the transformation of textual descriptions into visual representations of products. The work begins with an overview of the current state of technology in the field of genAI tools for text-to-image conversion, highlighting different approaches and techniques used for generating visualizations. In the next step, the characteristics of the most popular available genAI tools were analyzed, as well as their applicability in the context of product development, with an emphasis on creating detailed and accurate representations that reflect the functional and aesthetic characteristics of products. An important part of the research is the evaluation of factors that influence the quality of generated visualizations, particularly the structure and content of textual descriptions. Based on these findings, examples of best practices in creating visualizations were analyzed, including the development of guidelines for formulating effective prompts that encourage better results in image generation. The research includes a detailed study illustrating the application of genAI tools for visualizing different product concepts with specific requirements, showing how genAI tools can support innovative conceptualization and design processes, enabling rapid experimentation with different concepts and their iterative improvement. Finally, the proposed guidelines for creating visualizations were validated through experimental scenarios with users who have no experience using such genAI tools. This approach allowed for a quantitative analysis of the effectiveness of the guidelines in practice, confirming their practical applicability and ability to enhance the conceptualization and design process of products. The research results presented in this thesis provide insight into the potential of genAI technologies to transform the product development process, highlighting their importance in industrial, educational, and research contexts.
Mateo Šimac
2024
Master thesis
