GlyphCreator: Towards Example-Based Automatic Generation of Circular Glyphs

Lu Ying, Tan Tang, Yuzhe Luo, Lvkesheng Shen, Xiao Xie, Lingyun Yu, Yingcai Wu

View presentation:2021-10-29T13:15:00ZGMT-0600Change your timezone on the schedule page
2021-10-29T13:15:00Z
Exemplar figure, described by caption below
Circular glyphs in (a1) EnsembleLens [68], (b1) SmartAdP [35], (c1) VisMatcher [28], (d1) VAICo [51], (e1) SeqDynamics [65], (f1) Visual IVO Editor [38], (g1) DropoutSeer [8], (h1) MutualRanker [34]. (a2)-(h2) shows the circular glyphs generated by GlyphCreator based on (a1)-(h1), respectively. Dashed arrows are used to associate glyph components with corresponding categories.
Fast forward

Direct link to video on YouTube: https://youtu.be/fSRTmJnPaMo

Abstract

Circular glyphs are widely used in different fields because of their effectiveness in representing multidimensional data. However, the creation of circular glyphs remains a difficult task due to the demand for professional design skills and laborious design processes. This paper presents an interactive authoring tool called GlyphCreator to support the example-based generation of circular glyphs. Given an example circular glyph and multidimensional input data, GlyphCreator can promptly generate a list of design candidates and supports interactive editing on the candidates to satisfy different design requirements. To develop GlyphCreator, we first derive a design space of circular glyphs from summarizing the relation between different visual elements. With this design space, we build a circular glyph dataset and develop a deep learning model for glyph parsing. The model is able to deconstruct a circular glyph bitmap into a series of visual elements. Next, we propose an interface with effective interactions to help users bind the input data attributes to visual elements and customize visual styles. We evaluate the parsing model through a quantitative experiment and demonstrate the use of GlyphCreator through a usage scenario. The effectiveness of GlyphCreator is confirmed through user interviews.