This poster was presented at Computer Graphics & Visual Computing (CGVC24) at City St George's, University of London.
Aron Owen and Jonathan Roberts
Traditional visualisation designers often begin by sketching their ideas before implementation. With the advent of generative AI, these sketches can be transformed into AI-generated visualisations through carefully crafted prompts. However, guiding generative AI to create appropriate and effective visualisations can be challenging. We propose a novel design dialogue process where designers articulate their thought processes as they work and later convert these textual narratives into prompts for generative AI. This approach not only aids generative AI in generating accurate visualisations but also helps designers identify and emphasise critical aspects of their concepts, thereby enhancing the overall design process. By integrating human creativity with the capabilities of generative AI, this method fosters a more deliberate and effective visualisation design process. Moreover, it allows for rapid iteration and refinement, leading to higher quality and more innovative visualisations. With its potential to maximise the synergy between human creativity and generative AI capabilities, this method can inspire more creative and effective visualisation designs, making the design process more accessible and efficient across visualisation design and perhaps other disciplines.