EuroVisPosters2019

Permanent URI for this collection

https://diglib.eg.org/handle/10.2312/2632785

Browse

Now showing 1 - 20 of 27

Accurate and Memory-Efficient GPU Ray-Casting Algorithm for Volume Rendering Unstructured Grid Data
(The Eurographics Association, 2019) Gu, Gibeom; Kim, Duksu; Madeiras Pereira, João and Raidou, Renata Georgia
We present a novel GPU ray-casting algorithm for volume rendering unstructured grid data. We employ the per-pixel intersection list to guarantee accurate results for non-convex meshes. For efficient memory access for the lists on the GPU, we represent the intersection lists for all faces as an array. To increase ray-coherency in a thread block and improve memory access efficiency, we propose an image-tile based ray distribution method. Also, we found that a prior approach using a per-thread local buffer to reduce redundant computation is not proper for recent GPUs. Instead, we use an on-demand calculation strategy that achieves much better performance even when it allows duplicate computation. With a GPU, our method achieved up to 36.5 times higher performance for the ray-casting process, and 18.1 times higher performance for the entire volume rendering, compared with the Bunyk algorithm using a CPU core. Also, our approach showed up to 8.2 times higher performance than a GPU-based cell projection method while generating more accurate rendering results.
Comparative Analysis with Heightmaps in Virtual Reality Environments
(The Eurographics Association, 2019) Kraus, Matthias; Buchmüller, Juri; Schweitzer, Daniel; Keim, Daniel A.; Fuchs, Johannes; Madeiras Pereira, João and Raidou, Renata Georgia
3D heightmaps can be considered as an extension of heatmaps using the third dimension to encode the respective value by height, often in addition to encoding it by color. In contrast to 2D heatmaps, 3D heightmaps allow a superposition without aggregation. However, they also have the general disadvantages of 3D visualizations, such as occlusion and perceptual distortion. Previous research has revealed various advantages of stereoscopic displays and virtual reality (VR) in the context of 3D visualizations, for example, concerning memorization, depth perception, and collaboration. In this paper, we present a novel technique to compare heightmaps in VR by introducing a multi-layer approach of stacked heightmaps. We demonstrate the applicability and usefulness of our method by means of a use case on comparative crime data analysis.
CrossWidget: a D3.js Plugin to Support Multiple Scented Cross Filtering Activities
(The Eurographics Association, 2019) Blasilli, Graziano; Lenti, Simone; Palleschi, Alessia; Madeiras Pereira, João and Raidou, Renata Georgia
Filtering is one of the basic interaction techniques in information visualization. To better support a filtering activity, the visualization should be enriched by selectors composed of different visual elements and data distribution (i.e., scented widgets), in order to improve the navigation of information spaces and the comprehension of the filtering result. Thus, we propose CrossWidget, an open source JavaScript D3.js plugin, aimed at supporting cross filtering activities. To better perform multiple selections, to understand their result, and to better investigate the relationships among attributes of the dataset, this plugin proposes a set of ''multiple scented widgets'' that can be easily customized and integrated into Web visualization systems.
A Descriptive Framework for Stories of Algorithms
(The Eurographics Association, 2019) Liem, Johannes; Henkin, Rafael; Wood, Jo; Turkay, Cagatay; Madeiras Pereira, João and Raidou, Renata Georgia
Data-driven stories, widely used in journalism and scientific communication, match well with the recent focus on interpretable machine learning and AI explainability. Current technologies allow authors to break away from narratives that reflect traditional analytical workflows. To support designing such types of stories, we introduce a descriptive framework that helps identifying narrative patterns and other characteristics of algorithm-related stories. We describe the design space within the framework and demonstrate how to apply to an example of an algorithm-centered story, discussing potential future steps.
EuroVis 2019 Posters: Frontmatter
(The Eurographics Association, 2019) Madeiras Pereira, João; Raidou, Renata Georgia; Madeiras Pereira, João and Raidou, Renata Georgia
Improving the Scalability of Interactive Visualization Systems for Exploring Threaded Conversations
(The Eurographics Association, 2019) McNutt, Andrew M.; Kindlmann, Gordon L; Madeiras Pereira, João and Raidou, Renata Georgia
Large threaded conversations, such as those found on YCombinator's HackerNews, are typically presented in a way that shows individual comments clearly, but can obscure larger trends or patterns within the conversational corpus. Previous research has addressed this problem through graphical-overviews and NLP-generated summaries. These efforts have generally assumed a particular (and modest) data size, which limits their utility for large or deeply-nested conversations, and often require non-trivial offline processing time, which makes them impractical for day-to-day usage. We describe here Forum Explorer, a Chrome extension that combines and expands upon prior art through a collection of techniques that enable this type of representation to handle wider ranges of data in real time. Materials for this project are available at https://osf.io/nrhqw/.
Interactive 3D Visualizations of Laser Plasma Experiments on the Web and in VR
(The Eurographics Association, 2019) Danielová, Mariana; Janečka, Pavel; Grosz, Jakub; Holý, Aleš; Madeiras Pereira, João and Raidou, Renata Georgia
Lasers play essential role in many areas of both fundamental as well as applied research. Experiments in plasma physics that study complex interactions of intense laser beams with various forms of targets rely heavily not only on theory, but also on physics simulations. Such simulations, computed on a several thousand CPU core cluster, often trace the behaviour of billions of charged elementary particles over hundreds of time steps, generating up to terabytes of data. Visualization of the data is a challenging process that is mostly performed on GPU clusters using software such as ParaView [Aya15]. While this workflow has many advantages and provides advanced features for visualization and interpretation of the raw data, it is limited in the dissemination of the results on the web in interactive and VR mode. Our approach at ELI Beamlines, a European laser research facility near Prague, uses our web-based interactive 3D visualization framework for exploration of curated and optimized simulation datasets. This application, while still in early stage of development, runs in a regular web browser and utilizes VR mode to offer scientists completely new point of view of their simulations. Based on in-house developed WebGL 2.0 implementation, the framework not only renders the dataset on the GPU in real-time at 60fps, but also provides orthogonal views, textual and numeric information, alongside a GUI containing timeline animation controls and layer visibility management, with additional graphical elements based on D3.js for plotting animated graphs and legends.
Interactive Hierarchical Quote Extraction for Content Insights
(The Eurographics Association, 2019) Knittel, Johannes; Koch, Steffen; Ertl, Thomas; Madeiras Pereira, João and Raidou, Renata Georgia
This work presents a new approach to visually summarize large micro-document collections such as tweets. We extract frequent patterns of phrases as shortened quotes to present analysts an overview of popular snippets and statements, enabling more specific insights into large text collections compared to keyword-based visualizations. In our hierarchical structure, each quote can be the starting point to extract more fine-grained patterns on a subset of sentences that match the parent pattern. We show that our approach is scalable by applying it to millions of tweets.
LOOPS: Visual Analysis Tool for Non-Periodic Protein Structures
(The Eurographics Association, 2019) Ulbrich, Pavol; Horňák, Michal; Kozlíková, Barbora; Madeiras Pereira, João and Raidou, Renata Georgia
Protein molecules, the basic building blocks of all living organisms, are formed by chain sequences of amino acids. Due to internal forces, these chains are folded into a spatial representation with recurring patterns along the chain, called secondary structures. The basic secondary structures are a-helices and b-sheets, however, the protein function is also determined and influenced by other, non-periodic structural units, such as loops. Playing a pivotal role in the protein folding and dynamics, the loop classification and understanding may be necessary for modeling of shared structure ancestry (homology), protein structure prediction, and protein design. However, their non-repetitive characteristic denies the easy identification of their regularities. Therefore, we present LOOPS, a web-based tool for exploring and analyzing loops in proteins, that offers a high level of interaction with reference to other secondary structures. Displaying the additional information about a loop, such as its conformation, geometry, and multiple b-factors obtained from various methods, LOOPS provides a simple tool for analysis of loops in a protein sequence.
Model-Agnostic Visual Explanation of Machine Learning Models Based on Heat Map
(The Eurographics Association, 2019) Sawada, Shoko; Toyoda, Masashi; Madeiras Pereira, João and Raidou, Renata Georgia
It is essential to assess the trustworthiness of the machine learning models when deploying them to real-world applications, such as healthcare and risk management, in which domain experts need to make critical decisions. We propose a visual analysis method for supporting domain experts to understand and improve a given machine learning model based on a model-agnostic interpretable explanation technique. Our visualization method provides a heat map matrix as an overview of the model explanation and helps efficient feature engineering and data cleaning. We demonstrate our visualization method on a text classification task.
A Multi-Scale Animation Framework for Biological Fibrous Structures
(The Eurographics Association, 2019) Klein, Tobias; Viola, Ivan; Mindek, Peter; Madeiras Pereira, João and Raidou, Renata Georgia
Fibrous structures are ubiquitous in cell biology and play essential structural and functional roles in the life cycle of a cell. They are long polymers, such as DNA carrying genetic information, or filaments forming the cytoskeleton, crucial for cell division and maintaining the cell shape. In order to disseminate new findings of such structures to peers or a general audience, animated 3D models of these structures have to be created, as they are too small to be imaged with microscopes. However, this is a tedious task carried out by scientific animators, who manually create expressive visual representations of biological phenomena. In this work, we present a novel concept which simplifies the process of animating multi-scale procedural models of biological fibrous structures. In contrast with existing work in the domain of molecular visualization, our approach can also capture dynamics, which are important to show when communicating biological processes.
Real-Time Gaze Mapping in Virtual Environments
(The Eurographics Association, 2019) Kraus, Matthias; Kilian, Timon Urs; Fuchs, Johannes; Madeiras Pereira, João and Raidou, Renata Georgia
In order to analyze an analyst's behavior in an immersive environment, his or her eye movements can be monitored using eye trackers. Hereby, points of individual interest can be objectively identified, for instance, to assess the usability and intuitiveness of a framework. However, this technique can be used not only as a post-event analysis tool but also to assist an ongoing exploration of a virtual environment. With this poster, we present a technique that allows a real-time gaze map creation which supports the immersed analyst by providing real-time feedback on the user's own activity. In our approach, all surfaces in the virtual environment are enwrapped with a mesh structure. The grid structure recognizes when a user drifts with his or her eyes above it and increments weights of activated node points. This allows highlighting areas that have been observed, but also those that have not been observed - also when they are occluded by other objects or surfaces. We tested our technique in a preliminary qualitative expert study and received helpful feedback for further improvements.
Shapes of Time: Visualizing Set Changes Over Time in Cultural Heritage Collections
(The Eurographics Association, 2019) Salisu, Saminu; Mayr, Eva; Filipov, Velitchko Andreev; Leite, Roger A.; Miksch, Silvia; Windhager, Florian; Madeiras Pereira, João and Raidou, Renata Georgia
In cultural heritage collections, categorization is a central technique used to distinguish cultural movements, styles, or genres. For that end, objects are tagged with set-typed metadata and other information, such as time of origin. Visualizations can communicate how such sets organize a collection - and how they change over time. But existing interfaces fall short of a) representing an overview of temporal set-developments in an integrated fashion and b) of representing the set elements (i.e., the cultural objects) themselves to be contemplated on demand. Against this background, we introduce two integrated visualization techniques - a superimposition and a space-time cube view - depicting the development of sets and their elements over time. We share first results from a qualitative evaluation with casual users and outline open challenges for future research.
A Study on Reflecting User's Emotions Using a Painterly Rendering System
(The Eurographics Association, 2019) Lee, Junghyun; Seo, Sanghyun; Yoon, Kyunghyun; Madeiras Pereira, João and Raidou, Renata Georgia
Painters express visual experience in paintings through their own form of artistic expression and viewers feel various emotions about the elements appearing in paintings. However, it is very difficult for non-experts to express their emotions in painting by using painting elements as painters do. Thus, we aim to help users to express their feelings by improving their understanding of painting elements. To do this, we created a variety of paintings based on elements using a painterly rendering system with painting features. Then, the relationship between painting elements and emotions was examined through the study of users. We present a method of generating a painting expressing a specific emotion by applying elements according to certain emotional values.
Supporting Volumetric Data Visualization and Analysis by Combining Augmented Reality Visuals with Multi-Touch Input
(The Eurographics Association, 2019) Sereno, Mickael; Besançon, Lonni; Isenberg, Tobias; Madeiras Pereira, João and Raidou, Renata Georgia
We present our vision and steps toward implementing a collaborative 3D data analysis tool based on wearable Augmented Reality Head-Mounted Display (AR-HMD). We envision a hybrid environment which combines such AR-HMD devices with multi-touch devices to allow multiple collaborators to visualize and jointly discuss volumetric datasets. The multi-touch devices permit users to manipulate the datasets' states, either publicly or privately, while also proposing means for 2D input for, e. g., drawing annotations. The headsets allow each user to visualize the dataset in physically correct perspective stereoscopy, either in public or in their private space. The public space is viewed by all, with modifications shared in real-time. The private space allows each user to investigate the same dataset with their own preferences, for instance, with a different clipping range. The user can later decide to merge their private space with the public one or cancel the changes.
A Survey on Sleep Visualizations for Fitness Trackers
(The Eurographics Association, 2019) Aravind, Ranjini; Blascheck, Tanja; Isenberg, Petra; Madeiras Pereira, João and Raidou, Renata Georgia
We contribute the results of an exploratory study and a survey on visualizations for fitness trackers. Fitness trackers are becoming ubiquitous trackers of personal data. They often come with small attached displays that show micro visualizations of data such as heart rate, step counts, sleep duration, or number of floors climbed. Unfortunately, little is known about how wearers of fitness trackers use and perceive these micro visualizations. To collect data on the use of fitness visualizations, we conducted ten personal interviews with regular wearers of fitness trackers. Inspired by frequent responses regarding sleep tracking, we deployed an online questionnaire specifically on sleep visualizations for fitness trackers. Our results show that most participants were interested particularly in seeing previous night's sleep data on their fitness trackers and preferred visualizations that were easy to read like the hypnogram and bar as well as donut charts for sleep phases and duration.
A Systematic Review of Online Bitcoin Visualizations
(The Eurographics Association, 2019) Tovanich, Natkamon; HEULOT, Nicolas; Fekete, Jean-Daniel; Isenberg, Petra; Madeiras Pereira, João and Raidou, Renata Georgia
We contribute a systematic review of online visualizations of the Bitcoin blockchain. Bitcoin is currently the most active cryptocurrency with the largest market share among other cryptocurrencies. It has attracted a large user base and more and more businesses are beginning to accept Bitcoin as payment. While there are still relatively few visualization research papers on Bitcoin, a growing number of online tools visualize data about the Bitcoin blockchain. We provide a first systematic assessment of these online tools to inform future research efforts on making the Bitcoin blockchain more accessible.
Text Visualization Revisited: The State of the Field in 2019
(The Eurographics Association, 2019) Kucher, Kostiantyn; Kerren, Andreas; Madeiras Pereira, João and Raidou, Renata Georgia
Text and document data visualization is an important research field within information visualization and visual analytics with multiple application domains including digital humanities and social media, for instance. During the past five years, we have been collecting text visualization techniques described in peer-reviewed literature, categorizing them according to a detailed categorization schema, and providing the resulting manually curated collection in an online survey browser. In this poster paper, we present the updated results of analyses of this data set as of spring 2019. Compared to the recent surveys and meta-analyses that mainly focus on particular aspects and problems related to text visualization, our results provide an overview of the current state of the text visualization field and the respective research community in general.
Towards a WHAT-WHY-HOW Taxonomy of Trajectories in Visualization Research
(The Eurographics Association, 2019) Allain, Kevin; Turkay, Cagatay; dykes, jason; Madeiras Pereira, João and Raidou, Renata Georgia
Effective analysis of movement often requires a comprehensive approach where computational and visual methods are combined to address a wide variety of tasks involving movers with diverse characteristics. In order to help the process of designing effective methods for a wide range of movement analysis cases, we develop a provisional taxonomy that links what Brehmer et al. [BM13] term statements of WHY-WHAT-HOW with tasks, types of movers, context and methods used to compute or visualize data. Within this document we present the origin of this taxonomy, the process we followed to populate it, discuss the novel categories within it, and finally use it to explore relationships between elements of trajectory analysis. Our main contribution is to provide a new means of connecting elements of WHY-WHAT-HOW when analysing trajectories.
Visual Analysis of Aluminum Production Data with Tightly Linked Views
(The Eurographics Association, 2019) Jekic, Nikolina; Mutlu, Belgin; Faschang, Mario; Neubert, Steffen; Thalmann, Stefan; Schreck, Tobias; Madeiras Pereira, João and Raidou, Renata Georgia
Monitoring, analyzing and determining the production quality in a complex and long-running process such as in the aluminum production is a challenging task. We aim to support production data exploration in the aluminum industry. To this end, we developed the first version of the interactive visual analytics tool ADAM. The main aspect of concern is product quality, which is obtained from the quality inspection of aluminum plates at the end of the production process. A set of tightly linked views of production parameters with cross-filtering capability support the inspection of factors possibly influencing the product quality. ADAM allows highly responsive forward and backward search in the quality and production parameter space, leading to an understanding of important parameters, and supporting production planning and process improvement. Our approach was designed in an iterative development cycle guided by domain requirements from a major aluminum producer. We introduce the domain problem, propose a visual analytics design to support the problem, and demonstrate by application to real production data the usefulness and possible insights which can be obtained.

Browse

Browsing EuroVisPosters2019 by Title

Results Per Page

Sort Options