Visualisation Matters 2021 is designed to inspire audiences and demonstrate how art, science, design and engineering can be closely connected through visual creativity.
UNSW EPICentre and Woollahra Council invite you to participate in an Artwork-Hackathon as part of the Double Vision Public Art Project. It will be held in conjunction with Visualisation Matters 2021. Please register and submit your work!
Professor Claire Annesley (Dean, ADA) - Visualisation Matters Opening . Tomasz Bednarz - Program, Announcements and Acknowledgments
3 September 2021 9:00 am - 9:10 am
Keynote: Kerrie Mengerson - Connecting R to VR: From visualisation to elicitation to illumination
3 September 2021 9:10 am - 9:50 am
Kerrie Mengersen is a Distinguished Professor of Statistics at QUT, Deputy Director of the ARC Centre of Excellence in Mathematical and Statistical Frontiers, and Director of the QUT Centre for Data Science. Her research interests include the development of new –usually Bayesian –statistical and machine learning models and computational algorithms, with applications in health, environment and industry.
Connecting R to VR: From visualisation to elicitation to illumination VR and its derivatives have become standard visualisation and animation tools for scientists, and this is quickly extending to data science. However, most data scientists find that there is still a gap between the development of the VR environment and the analysis of the derived information. In this presentation, I will introducetheR2VR software (R to Virtual Reality)that has been built by our team, led by MPhil student Jon Peppinck under the co-supervision of Dr Julie Vercelloni. R2VR allows for VR scenes to be created and updated in R, and allows citizens to annotate virtual reality imagery in both 2D and 3D, with an interoperable flow of data between the VR and R. The software is demonstrated through a case study involving the classification of coral cover in The Great Barrier Reef.
Session: Liminalities of Visualisation
3 September 2021 9:50 am - 10:50 am
Guandong Xu - Deep Visual Analytics and its Applications: Nowadays, the uses of massive amounts of data are rapidly increasing in many applications such as (i) healthcare analysis, (ii) fintech ecosystem, and (iii) tourism management, etc. The analysis of these data are grimy, irreconcilable and complex. As a result, vast amounts of time and money are often lost. To address these data issues, new advancements such as Deep Visual Analytics (DVA) has been received significant attention for solving various complex problems. Usually, it consists of two major parts such as 1) data visualization which is an emerging field in the current situation, 2) deep learning which adds more insights, excels at knowledge communication, and discovering strategies by applying encoding techniques to transfer abstract data into meaningful representation. It fosters the worthwhile assessment, correction and rapid improvement of big data with meaningful interactive visualization (IV). Finally, it views as the science of analytical reasoning empowered by interactive visualizations.
Josh Harle - John will discuss ethics around creating virtual reality experiences, specifically mining and military applications. While virtual reality has been broadly embraced as “the ultimate empathy machine”, a robust argument can be made for virtual reality operating rather to desensitise participants to emotional responses. Examples will be given of VR for immersion treatment for phobias, exposure therapy for addiction, habituating muscle memory, and HR training experiences for effectively firing workers. Military VR will be explored further in the context of desensitisation and the historical development of military training from Dave Grossman’s On Killing. Finally, looking at the use of Virtual Reality for journalism, it will acknowledge that the novelty of VR has exposed (particularly affluent) audiences to particular issues, explore the ethical implications of desensitisation and compassion fatigue of the audience to the subject, and a brief discussion of possible approaches to avoid this.
Liz Gilleran - Why You Should’ve Started Blogging Yesterday: Many of our best ideas originate in our heads, but unfortunately that’s where many of them stay. Maybe we’re saving it for a paper, maybe for a presentation, but ultimately we’re limiting our own audience. Blogging and putting our ideas on the web is the way to unlock bigger audiences and get them interested in our research, ideas, and methods. In this talk I will detail how I went from PhD dropout to blogging about my professional practice. I will detail why I set up my blog the way I do, the pitfalls and benefits of publishing on other people’s publications, and the benefits of opening up online. Everything from measuring the success and impact, to overcoming the confidence hump and what pushes me to keep going even when I don’t want to.
Philip Gough - The Flow of Visualisation: what can non-expert users get out of creative data visualisation: Data visualisation serves to provide insight, but for the general public, that may not be enough to engage an audience, who may be busy, distracted, or in any range of contexts where “insight” is not their desired outcome. In this talk, Dr Phil Gough from the Sydney School of Architecture Design and Planning will discuss possibilities for the user experience beyond insight. We will see how both positive psychology and biology can be leveraged in visualisation design to help engage an audience with data and provide an interesting user experience.
Rowan Hughes - Real-Time Crowd Simulation as a Design Tool & Getting Over the Data Blues
Combining real-time simulation into the visual analytics pipeline can add a powerful tool to a designer’s toolbox. This presentation introduces an architectural design framework that incorporates crowd simulation directly into the design process. This allows designers to understand how their designed space operates in practice and facilitates real-time iteration over their designs to improve throughput, egress speed, space usage, etc. Aside from floorplan design, designers can also factor in crowd behaviour and cultural factors by adjusting model parameters. In current approaches, agents are often represented with 2D proxies that move about a plane. Our framework allows designers to seamlessly transition to the space in 3D or work with another user and engage in real-time 3D co-design. We hope this will enable more precise human behaviour communication that helps designers design solutions more holistically.
Morning Coffee Break
3 September 2021 10:50 am - 11:10 am
Session: Immersed in Vis
3 September 2021 11:10 am - 12:10 pm
Tim Dwyer - Immersive Analytics in a Connected World: We reflect on research from the Monash University Data Visualisation and Immersive Analytics lab since 2005, and with our colleagues elsewhere, which spans network visualisation, human-in-the-loop optimisation and immersive analytics. Network visualisation has the potential to help people to better understand the highly connected and complex world in which they live. As researchers our group are both users of optimisation technologies, to create visualisations that are in some sense optimal, and also contributors to those technologies, seeking to create human-centred optimisation tools for many applications. This leads to a discussion of our more recent work in the area of Immersive Analytics, which aims to bring data out of computer centres and into the world around us.
Maxime Cordeil - Embodied Data Interaction: Visualising data is fundamental to understand complex phenomena and patterns, predict future trends and eventually make informed decisions. The data visualisation process requires human input and therefore significant amounts of interaction are required to support the effective exploration of data (e.g. selecting data, filtering, specifying views etc). Interaction with data and their visual representation has largely been designed and developed for 2D screen/mouse/ keyboards devices, which creates a barrier between the data and people who seek to understand them. Modern VR/AR technology allows for more natural, 3D rich interaction and presents unique opportunities to redesign the data interaction pipeline to visually explore data in immersive environments to enhance data understanding. In this talk, I will present my work on embodied data interaction in mixed and virtual reality. I will present the emerging design space of spatial interaction with complex data and present some exemplar work that aim at reducing the gap between people and their data.
Haifeng Shen - Human-AI Interactive and Continuous Sensemaking through Annotated Attention Heatmaps: Advances in AI, especially the stunning achievements of Deep Learning (DL) in recent years, have shown DL models possess remarkable understanding towards the logic reasoning behind the solved tasks. However, human understanding towards what knowledge is captured by deep neural networks is still elementary and this has a detrimental effect on human’s trust in the decisions made by AI systems. While progress has been made to elucidate the reasoning processes of AI algorithms in such a way that makes sense to humans through Explainable AI, humans’ feedback on AI explanations is generally lacking. This is a critical gap because if humans were able to point out incorrect reasonings in a way that makes sense to AI algorithms, they would be able to fine tune themselves for performance improvement by incorporating humans’ feedback. In this talk, we present a human-AI interactive and continuous sensemaking framework that supports effective human-AI interaction by unveiling the dual black boxes to each other for mutual sensemaking through annotated attention heatmaps. A case study of image classification using real-world fracture X-rays has shown significant improvement of classification accuracy with the framework.
Quang Vinh Nguyen - Enabling Genomic Data Analytics with Virtual Reality: Virtual Reality (VR) environment has been gradually gained traction in data analytics thanks to its less distraction, more space and natural interactions integration. VR technologies could provide significant improvement on scientific genomic data visualisation and interpretation, especially when merging with machine learning models. This presentation presents a milestone in how VR has been used in genomic data analytics. In the project, we learn from our existing research works on visualisations and machine learning and extract essential requirements based on the feedback from the end-users such as clinicians and researchers to enhance genomic data analytics in virtual and mixed reality environments.
Bakir Babic - Immersive visualisation for a large metrological data set: A purpose of scientific visualisation is to facilitate the transfer of factual information about underlying data. Visualisation of a large data set requires an appropriate visualisation method where instead of focusing on expected features all acquired data are treated on an equal footing. Here, we identify critical differences between raw and processed data acquired from displacement measurements at the nanoscale. We illustrate the strength of immersive data visualisation in rapidly and intuitively identifying non-trivial data trends, supported by correlation with the associated measurement uncertainties. These insights can be directly leveraged for optimisation of instrumentation and measurement protocols, application of appropriate measurement corrections, and iterative experimental design changes.
3 September 2021 12:10 pm - 1:20 pm
Session: Beyond Visualisation
3 September 2021 1:20 pm - 2:10 pm
Khyiah Angel - AR for young adult literature: The nature of reading and writing has changed. Young adults tend to preference the visual and multimodal above the textual and, as such, creators need to find ways of thinking beyond the traditional tome to search for more relevant ways to engage young people with text. The visual has long supplanted the linguistic text-based approach to reading and writing, and as new technologies continue to emerge, so too, do new ways of creating and consuming texts. Novel reading still occurs but in much more fluid and adaptive, and diverse ways. This work examines the role and use of Augmented Reality to breathe life into long-form fiction novels by drawing on the visual and combining image, video, audio and effects to create an ‘almost traditional’ social-semiotic multimodal reading experience.
Cameron Edmond - No more heroes: Narrative structures for data stories Despite the field’s a long history, formal narrative structures for narrative visualisation/data storytelling are yet to emerge. While the data visualisation community makes leaps and bounds in terms of data visuals and interactivity, we continue to grapple with even the basic building blocks of narrative. It is not uncommon for narrative visualisation strategies to engage with of Joseph Campbell and similar writers, introducing problematic, colonial baggage. Pragmatically, carrying over structures designed for fiction into data communication provides its own ethical quandaries. We offer a novel “data-focused” narrative structure and tool that rejects traditional narrative structures. Our solution borrows from computational linguistics and digital humanities, using Rhetorical Structure Theory (RST) – a model for analysing and generating text – to create a narrative form. This presentation will introduce our modified RST model and the tool that drives it, designed to empower narrative visualisation authors to craft inclusive, data-focused narratives.
Naseem Ahmadpour - Make the talk: affective storytelling in design research: The role of narrative storytelling has been established and appreciated in human-centred design research, particularly in capturing the affective experiences of users. However, it can be challenging for people to communicate stories and articulate experiences they have never considered before. In this talk, I will provide examples of visual storytelling tools and techniques I have used to encourage research participants to engage in storytelling through ‘making’. This approach acknowledges the complexities of externalising narratives of our past and dreams of the future and empowers individuals to shape the direction of open-ended design inquiries. Notions of care, responsibility and agency are integral to this approach. I will end the talk by outlining the design opportunities that may surface through following this approach.
Tony Huang - Making graph visualization a user-centered process: It is commonly agreed that graphs that are drawn based on some pre-defined quality metrics, or aesthetics are more likely to be not only beautiful but also comprehensible. However, those aesthetics were originally proposed by researchers based on their intuitions from a designer’s perspective. Although some of them have been empirically validated, to produce truly user cognitively effective visualizations, well-grounded cognitive theories and design guidelines are needed. Attempts have been made to explore a cognitive approach that follows a user-centered process in visualizing graphs. In this talk, I will introduce theories from other disciplines, summarize methodologies that have been used in prior research, present case studies and discuss their implications for producing user-centered visualizations.
Afternoon Tea Break
3 September 2021 2:10 pm - 2:30 pm
Keynote: John McGhee - Nanoscape: The Art and Science of building 3D real-time interactive virtual cell environments
3 September 2021 2:30 pm - 3:10 pm
Associate Professor John McGhee is the Director of the 3D Visualisation Aesthetics Lab. John’s academic research work explores arts-led and design-led modes of visualising complex scientific and biomedical data. His research investigates the application of creative practice, technical innovation and immersive platforms to new data visualisation challenges. Areas of research include virtual world design, scientific discovery, education and engagement, biomedical communication and clinical interaction. This has culminated in John being recognised as one of UNSW Australia’s 21 Rising Stars in research leadership.
Nanoscape: The Art and Science of building 3D real-time interactive virtual cell environments: Our understanding of cellular and structural biology has reached unprecedented levels of detail, and computer visualisation techniques can be used to create three-dimensional (3D) representations of cells and their environment that are useful in both teaching and research. However, extracting and integrating the relevant scientific data, and then presenting them in an effective way, can pose substantial computational and aesthetic challenges. In this keynote presentation I explore how computer artists, experts in computer graphics and cell biologists have collaborated to produce a tool we call Nanoscape that allows users to explore and interact with 3D representations of cells and their environment that are both scientifically accurate and visually engaging.
Session: Visualising Humanity
3 September 2021 3:10 pm - 4:00 pm
Can Yalçınkaya - Comics as Research: Creating autoethno(graphic) narratives and the scholar as a cartoon character: As a graphic narrative art form, comics construct meaning on two distinct but interwoven levels – images and words. In the last few decades, comics have become an increasingly popular medium to communicate journalistic reporting, research findings, issues of political significance, and other non-fiction genres. Their power to show, in pictures and other visual symbols, what words may not be able to describe puts them in a unique position to convey meaning. Comics are arguably the perfect medium with which to share autoethnographic and autobiographical research, due to their emphasis on individual cartooning styles and the explicit subjectivity of the world(s) they portray, which makes the constructed nature of non-fiction/research-based writing more transparent, challenging the illusion of objectivity that often accompanies these texts. In this talk, I will discuss the affordances and limits of graphic research, based on my experience as a creative practitioner and academic using comics and zines to disseminate my research findings.
John Lock - Seeing what the AI is thinking: visual analytics explaining AI models in biomedical imaging: Biomedical imaging is producing a tidal wave of data with potential to transform human health. Yet in scale and complexity this data is beyond the scope of human analysis. Only artificial intelligence (AI) – particularly Deep Learning – can fully leverage this information for human betterment. But, how can we trust our health to AI models that we cannot comprehend? Overcoming this “explainability gap” is vital for uptake of AI models that can literally save lives. We introduce “BioDive”, a visual analytics platform developed in parallel with a novel Deep Learning architecture to immerse biomedical researchers in 3D representations of “what the AI is thinking”. Coupling biomedical images into AI-learned data representations enables comparison of AI interpretations versus our own visual understanding. Supporting human feedback to improve AI models, this fusion of visual analytics and AI begins to bridge the explainability gap to empower health advances based on biomedical image analysis.
Deborah Richards - Intelligent Virtual Agents: who cares what they look like and why?: The appearance of intelligent virtual agents has been shown to impact the way people respond to the character. In helping contexts, building human-agent rapport is important for a successful supportive relationship. Positive first impressions will get the relationship off to a good start and are typically based on appearance and preferences or biases. Together with a review of other studies, this talk will present the results of a study we conducted with 174 young adults (aged 18-30) to determine their preferences for the look of a character in a helping role with respect to the character’s age, gender and ethnic appearance. After briefly meeting and ranking 12 animated characters, participants interacted with their top-ranked character. We measured the level of rapport established and general attitude to using virtual humans for support. I will raise related ethical issues and the appropriate uses of this technology.
Humphrey Obie - Gravity++: An Alternative Graph-based Approach for Creating Interactive Visualisation Narratives: Interactive visualizations play a key part in the exploration and analysis of data, and the creation of visual data stories. We describe an alternative graph-based framework for developing interactive visualizations for the creation of coherent visual data stories. We realised our framework in a prototype tool named Gravity++. Gravity++ uses a graph architecture for modelling interaction, data navigation, and changes in visual representation. The combination of these graph models provides better support and flexibility for designing interactive visualizations, data navigation, and visual data stories. Our alternative graph-based framework aims to support the understanding and sense-making of the visual data story creation process.
Maria Lacey - Double Vision
3 September 2021 4:00 pm - 4:30 pm
Albert Einstein (1879 – 1955)
For more info or submit proposal of a lightning talk please e-mail epicentre(at)unsw(dot)edu(dot)au