1: Familiarity vs. Correctness: Arcs and Angles in Sunburst Charts [PDF]

Abstract: A Sunburst chart visualizes hierarchical data as concentric rings. Similar to a pie chart, a Sunburst chart may be read through arc lengths or sector angles, but its multilevel nature makes arc lengths across levels unreliable as a visual variable. In this study, we designed Sunburst chart variations to test whether they affect chart reading performance. We found that the participants used arc lengths to read chart values, confirming a previous study. Other factors in the experiment suggested that the values in focus should be placed in adjacent layers and in quadrant III and IV as the participants tended to correctly read data values through angles instead.

2: Exploring Handmade Soap as a Visualization Medium [PDF]

Abstract: We ordered handmade soaps to visually reflect 12 digital visualization designs in our previous work. The soap colors precisely matched the designs while the shapes varied. Our expectations of the 12 designs did not match the soap maker's evaluations. Their comments were noted for picking designs suitable for soap making in the future. This can be helpful for further explorations of other physical materials for visualization, especially in every objects.

3: Visualization Transfer from 2D Image to 3D Volume [PDF] [Video] [Best]

Abstract: Traditional visualization work focuses on collaborative visualization and 2-D data analysis, which is limited when analyzing complex 3-D data. In 3-D volume data analysis, visualization transforms the invisible phenomena in volume data into visible, revealing the rich connotations contained in volume data. It enhances domain experts’ awareness of three-dimensional space. However, it is complicated for experts to mark target objects in 3-D images, because it is very difficult to select the target point in the 3-D image accurately. If all seed points can be automatically recommended to the 2-D image, and after manual selection, the required structure can be marked in the 3-D image by the seed point tracking algorithm, then domain experts will be able to greatly improve the efficiency of data interpretation. In this paper, we propose an interactive visualization system based on client-server seed point recommendation and tracking, which consists of a 3-D volume explorer and a 2-D slice analyzer, to help domain experts fully utilize their background domain knowledge to illustrate different parts of the data. The system realizes multi-user interaction and can run on multi-terminal devices because the client-side calculation is very small. It allows taking full advantage of the hardware resources because all the computation-intensive tasks especially for the whole data rendering can be allocated to the powerful server while the light-weight tasks can be allocated to the portable clients. For example, the brain surgeon expert, pulmonologist, and cardiologist can visualize and analyze the different subsets of the volumetric scientific data, that is, the corresponding subvolumes of the brain, heart, lungs, and blood vessels. We calculate the image entropy of all slices on the server side, and select the slice with the largest entropy, and then recommend all seed points according to the seed point recommendation algorithm based on meanshift. In addition, we also design a seed point tracking algorithm based on continuous scale space theory, which realizes the migration from 2-D to 3-D.

4: A Preliminary Model for the Design of Music Visualizations [PDF]

Abstract: Music Visualization is basically the transformation of data from the aural to the visual space. There are a variety of music visualizations, across applications, present on the web. Models of Visualization include conceptual frameworks helpful for designing, understanding and making sense of visualizations. In this paper, we propose a preliminary model for Music Visualization. We build the model by using two conceptual pivots – Visualization Stimulus and Data Property. To demonstrate the utility of the model we deconstruct and design visualizations with toy examples using the model and finally conclude by proposing further applications of and future work on our proposed model.

5: Visual Analytics for Autonomous Driving

Abstract: Autonomous driving has developed rapidly with artificial intelligence algorithms in perception, planning, controlling, and decision-making models. However, most autonomous driving technologies can not sufficiently support autonomous driving algorithms' understanding and optimization. Therefore, we propose a system based on data visual analytics methods through visual scenes and interactive exploration methods. The system provides a set of interpretable model environments integrating the spatiotemporal scenes and results of detection. It enables the autonomous driving model developers to perceive, understand, tune and verify the algorithm.

6: TopicOcean: Pair-wise Visual Comparison for Evolving Topics and Contexts in Social Media

Abstract: How do various topics trend rise and fall when they are spreading on social media? How does the composition of contextual words evolve? How to measure the relationship between topics and key actors? In this study, we propose TopicOcean, a timeline visualization design involving interactive selection and dynamic temporal aggregation that allow an intuitive understanding of the topics and contexts in two streams. The visual design reflects thematic and pluralistic aspects of social media information transmission through composing ocean bubbles and tree-style metaphors. Ocean bubbles represent keywords showing the increase and decrease of topics and contexts, and bilateral trees sequentially connect contexts into patterns. We introduce a correlation degree to arrange bubbles sequentially to weigh up the pair-wise relationship with key actors. Integrated with visual exploration, we propose algorithms for seeking pattern and spanning tree structure. To validate the effectiveness of the visual analysis techniques, we conducted a case study based on people's reaction to the presidential debate and verified the analysis by combining it with real media news.

7: Interactive Spiral: A Radial Visualization for Mobile Devices [PDF] [Video]

Abstract: We propose an interactive radial visualization of tabular data optimized for mobile devices. Our visualization employs a spiral layout, making it suitable for small handheld touchscreens. Users can view an overview of the data, search for specific data, and access details as needed. We hope this visualization technique will encourage users to interact with data rather than just looking at them passively.

8: Temporal Hierarchical Visualization of COVID-19 Epidemic Data [PDF]

Abstract: In this work, we present two novel visualizations for COVID-19 visual analysis, to deal with the hierarchical relationship in temporal data. By designing a novel tree visualization, we are able to give an overview of the temporal traits as well as spreading relationship over the epidemic transmission. To compensate for a detailed analysis, This work designs a visualization using the metaphor of musical notation, for an easy temporal analysis within a hierarchy. We collected the daily-reported COVID-19 cases in Shenzhen and show the usage of these two visualizations.

9: Smart Pick-Paste: Infographics Design from Examples [PDF]

Abstract: Existing infographics are usually considered as design resources to get inspired, or even use for infographics creation. However, most of infographics, in the form of images, are unstructured and hard for a direct reuse. In this work, we propose a novel interactive system with smart pick-and-paste of graphic elements, to help designers create infographics from existing infographic images. By parsing the underlying design patterns of infographics and automatically layering into a hierarchical structure, with a single scribbled sketch, the system supports users to efficiently pick elements by inferring users' intention of selecting groups. A series of paste-up operations across groups support efficient re-editing and generation of harmonized infographics.

10: Embedding Dynamic Information in Static Visualization Images with Residual Modeling [PDF]

Abstract: Embedding dynamic information in static visualization images is a practical but challenging problem. It can be used in various applications, such as provide extra data and interactivity through static visualizations. However, embedding large-scale information in a static image may lead to a high decoding error rate. We present a novel framework to implicitly embed dynamic information into a static visualization image while preserving the visual quality of the encoded image. The designed model considers the temporal redundancy within consecutive dynamic information and uses residual modeling in the encoding stage. The decoder network allows users to obtain the dynamic data with a static encoded visualization image. We demonstrate the expressiveness of our model through actual scenarios.

11: Interactive Visualization for Exploring Information Fragments in Software Repositories [PDF] [Video]

Abstract: Software developers explore and inspect software repository data to obtain detailed information archived in the development history. However, developers who are not acquainted with the development context suffer from delving into the repositories with a handful of information; they have difficulty discovering and expanding information fragments considering the topological and sequential multi-dimensional structure of repositories. We introduce ExIF, an interactive visualization for exploring information fragments in software repositories. ExIF helps users discover new information fragments within clusters or topological neighbors and identify revisions incorporating user-collected fragments.

12: Visual Storytelling for Science Popularization and Education: A Design Study of the Evolutionary Game Theory

Abstract: More and more attention has focused on data visualization for its ability of revealing stories within data. Particularly, visual storytelling can be applied as an efficient technique for science popularization and education. However, there is a lack of design space and implications for science popularization with visual storytelling of mathematical models, indicating a need to learn about how to design a visual storytelling for disseminating mathematical knowledge. Therefore, in this work, we summarize the design space for storytelling of mathematical models. Following the design space, we then propose a visual storytelling design for the evolutionary game theory.