2019 IEEE Visualization Conference (VIS)最新文献

{"title":"A Deep Learning Approach to Selecting Representative Time Steps for Time-Varying Multivariate Data","authors":"William P. Porter, Yunhao Xing, Blaise R. von Ohlen, Jun Han, Chaoli Wang","doi":"10.1109/VISUAL.2019.8933759","DOIUrl":"https://doi.org/10.1109/VISUAL.2019.8933759","url":null,"abstract":"We present a deep learning approach that selects representative time steps from a given time-varying multivariate data set. Our solution leverages an autoencoder that implicitly learns feature descriptors of each individual volume in a latent space. These feature descriptors are used to reconstruct respective volumes for error estimation during network training. We then perform dimensionality reduction of these feature descriptors and select representative time steps in the projected space. Unlike previous approaches, our solution can handle time-varying multivariate data sets where the multivariate features can be learned using a multichannel input to the autoencoder. We demonstrate the effectiveness of our approach using several time-varying multivariate data sets and compare our selection results with those generated using an information-theoretic approach.","PeriodicalId":192801,"journal":{"name":"2019 IEEE Visualization Conference (VIS)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130666161","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Point Movement in a DSL for Higher-Order FEM Visualization","authors":"Teodoro Collin, Charisee Chiw, L. R. Scott, John H. Reppy, G. Kindlmann","doi":"10.1109/VISUAL.2019.8933623","DOIUrl":"https://doi.org/10.1109/VISUAL.2019.8933623","url":null,"abstract":"Scientific visualization tools tend to be flexible in some ways (e.g., for exploring isovalues) while restricted in other ways, such as working only on regular grids, or only on unstructured meshes (as used in the finite element method, FEM). Our work seeks to expose the common structure of visualization methods, apart from the specifics of how the fields being visualized are formed. Recognizing that previous approaches to FEM visualization depend on efficiently updating computed positions within a mesh, we took an existing visualization domain-specific language, and added a mesh position type and associated arithmetic operators. These are orthogonal to the visualization method itself, so existing programs for visualizing regular grid data work, with minimal changes, on higher-order FEM data. We reproduce the efficiency gains of an earlier guided search method of mesh position update for computing streamlines, and we demonstrate a novel ability to uniformly sample ridge surfaces of higher-order FEM solutions defined on curved meshes.","PeriodicalId":192801,"journal":{"name":"2019 IEEE Visualization Conference (VIS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130844018","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Toward Perception-Based Evaluation of Clustering Techniques for Visual Analytics","authors":"Michaël Aupetit, M. Sedlmair, M. Abbas, Abdelkader Baggag, H. Bensmail","doi":"10.1109/VISUAL.2019.8933620","DOIUrl":"https://doi.org/10.1109/VISUAL.2019.8933620","url":null,"abstract":"Automatic clustering techniques play a central role in Visual Analytics by helping analysts to discover interesting patterns in high-dimensional data. Evaluating these clustering techniques, however, is difficult due to the lack of universal ground truth. Instead, clustering approaches are usually evaluated based on a subjective visual judgment of low-dimensional scatterplots of different datasets. As clustering is an inherent human-in-the-loop task, we propose a more systematic way of evaluating clustering algorithms based on quantification of human perception of clusters in 2D scatterplots. The core question we are asking is in how far existing clustering techniques align with clusters perceived by humans. To do so, we build on a dataset from a previous study [1], in which 34 human subjects la-beled 1000 synthetic scatterplots in terms of whether they could see one or more than one cluster. Here, we use this dataset to benchmark state-of-the-art clustering techniques in terms of how far they agree with these human judgments. More specifically, we assess 1437 variants of K-means, Gaussian Mixture Models, CLIQUE, DBSCAN, and Agglomerative Clustering techniques on these benchmarks data. We get unexpected results. For instance, CLIQUE and DBSCAN are at best in slight agreement on this basic cluster counting task, while model-agnostic Agglomerative clustering can be up to a substantial agreement with human subjects depending on the variants. We discuss how to extend this perception-based clustering benchmark approach, and how it could lead to the design of perception-based clustering techniques that would better support more trustworthy and explainable models of cluster patterns.","PeriodicalId":192801,"journal":{"name":"2019 IEEE Visualization Conference (VIS)","volume":"871 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132098817","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"EasyPZ.js: Interaction Binding for Pan and Zoom Visualizations","authors":"Michail Schwab, J. Tompkin, Jeff Huang, M. Borkin","doi":"10.1109/VISUAL.2019.8933747","DOIUrl":"https://doi.org/10.1109/VISUAL.2019.8933747","url":null,"abstract":"The creation of data visualizations has become easier as the skillbarrier to our tools has decreased. However, adding interactivity, such as gestures for pan and zoom, still requires significant coding expertise. We introduce an open-source library—EasyPZ.js—for the creation of multi-scale (pan and zoom) visualizations across desktop and mobile devices. EasyPZ is fully customizable and extendable with flexible options for interaction design. For example, it is easy to choose gestures which are compatible with selection interactions such as clicking. EasyPZ can be enabled on any SVG-based visualization on the web with one line of code, or by simply clicking a bookmark without requiring commitment to code changes. With this library, we contribute ways for the visualization community to more easily author interactive multi-scale visualizations.","PeriodicalId":192801,"journal":{"name":"2019 IEEE Visualization Conference (VIS)","volume":"67 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114139705","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Unsteady Flow Visualization via Physics Based Pathline Exploration","authors":"Duong B. Nguyen, Lei Zhang, R. Laramee, D. Thompson, R. O. Monico, Guoning Chen","doi":"10.1109/VISUAL.2019.8933578","DOIUrl":"https://doi.org/10.1109/VISUAL.2019.8933578","url":null,"abstract":"This work proposes to analyze the time-dependent characteristics of the physical attributes measured along pathlines derived from unsteady flows, which can be represented as a series of time activity curves (TAC). A new TAC-based unsteady flow visualization and analysis framework is proposed. The center of this framework is a new event-based distance metric (EDM) that compares the similarity of two TACs, from which a new spatio-temporal, hierarchical clustering of pathlines based on their physical attributes and an attribute-based pathline exploration are proposed. These techniques are integrated into a visual analytics system, which has been applied to a number of unsteady flow in 2D and 3D to demonstrate its utility.","PeriodicalId":192801,"journal":{"name":"2019 IEEE Visualization Conference (VIS)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116355079","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Interactive Dendritic Spine Analysis Based on 3D Morphological Features","authors":"Junyoung Choi, Sang-Eun Lee, Eunji Cho, Yutaro Kashiwagi, S. Okabe, Sunghoe Chang, W. Jeong","doi":"10.1109/VISUAL.2019.8933795","DOIUrl":"https://doi.org/10.1109/VISUAL.2019.8933795","url":null,"abstract":"Dendritic spines are submicron scale protrusions on neuronal dendrites that form the postsynaptic sites of excitatory neuronal inputs. The morphological changes of dendritic spines reflect alterations in physiological conditions and are further indicators of various neuropsychiatric conditions. However, due to the highly dynamic and heterogeneous nature of spines, accurate measurement and object analysis of spine morphology is a major challenge in neuroscience research. Here, we propose an interactive 3D dendritic spine analysis system that displays 3D rendering of spines and plots the high-dimensional features extracted from the 3D mesh of spines in three graph types (parallel coordinate plot, radar plot, and 2D scatter plot with t-Distributed Stochastic Neighbor Embedding). With this system, analysts can effectively explore and analyze the dendritic spine in a 3D manner with high-dimensional features. For the system, we have constructed a set of morphological high-dimensional features from the 3D mesh of dendritic spines.","PeriodicalId":192801,"journal":{"name":"2019 IEEE Visualization Conference (VIS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128736523","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"High Fidelity Visualization of Large Scale Digitally Reconstructed Brain Circuitry with Signed Distance Functions","authors":"Jonas Karlsson, M. Abdellah, Sébastien Speierer, A. Foni, Samuel Lapere, F. Schürmann","doi":"10.1109/VISUAL.2019.8933693","DOIUrl":"https://doi.org/10.1109/VISUAL.2019.8933693","url":null,"abstract":"We explore a first proof-of-concept application for visualizing large scale digitally reconstructed brain circuitry using signed distance functions. The significance of our method is demonstrated in comparison with using implicit geometry that is limited to provide the natural look of neurons or explicit geometry that requires huge amounts of memory and has limited scalability with larger circuits.","PeriodicalId":192801,"journal":{"name":"2019 IEEE Visualization Conference (VIS)","volume":"91 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124620241","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Towards Enhancing RadViz Analysis and Interpretation","authors":"M. Angelini, G. Blasilli, S. Lenti, A. Palleschi, G. Santucci","doi":"10.1109/VISUAL.2019.8933775","DOIUrl":"https://doi.org/10.1109/VISUAL.2019.8933775","url":null,"abstract":"RadViz plots are commonly used to represent multidimensional data because they use the familiar notion of 2D points for encoding data elements, displaying the original data dimensions that act as springs for setting the x and y coordinates. However, this intuitive approach implies several drawbacks and produces misleading visualizations that can confuse the user, even while analyzing a single data point. The paper attacks this problem following the well known idea of changing the order of the dimensions and introducing ancillary visualizations to mitigate some of RadViz drawbacks. In particular, the paper defines the notion of point optimal disposition of the dimensions for a single data point, generalizes this concept to a set of data points, and proposes effective heuristics for dealing with the intractable problem of exploring all the $frac{{left( {n - 1} right)!}}{2}$ dispositions of the dimensions along the RadViz circumference. Additional views, visual quality metrics, and a circular grid superimposed on the RadViz complement the attribute reordering strategy and provide a better understanding of the actual plot of the data elements.","PeriodicalId":192801,"journal":{"name":"2019 IEEE Visualization Conference (VIS)","volume":"77 7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130666770","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Towards Quantifying Multiple View Layouts in Visualisation as Seen from Research Publications","authors":"H. Al-Maneea, Jonathan C. Roberts","doi":"10.1109/VISUAL.2019.8933655","DOIUrl":"https://doi.org/10.1109/VISUAL.2019.8933655","url":null,"abstract":"We present initial results of a quantitative analysis of how developers layout the visualisations in their multiple view systems. Many developers create multiple view systems and the technique is commonly used by the visualisation community. Each visualisation shows data in a different way, and often user interaction is coordinated between the views. But it is not always clear to know how many views a developer should use, or what would be the best layout. We extract images of visualisation tools, across TVCG journal, conference, posters and workshop papers 2012-2018 to analyse the quantity and layout of the views in these visualisation systems. Focusing on view juxtaposition, we code the layout of 491 images and analyse view topology in juxtaposed views. Our analysis acts as a starting point to help designers create better visualisations, acts as a taxonomy of visualisation layouts, and provides a quantitative analysis of how many views developers have used in their visualisation systems.","PeriodicalId":192801,"journal":{"name":"2019 IEEE Visualization Conference (VIS)","volume":"49 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123750994","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Uncovering Data Landscapes through Data Reconnaissance and Task Wrangling","authors":"Anamaria Crisan, T. Munzner","doi":"10.1109/VISUAL.2019.8933542","DOIUrl":"https://doi.org/10.1109/VISUAL.2019.8933542","url":null,"abstract":"Domain experts are inundated with new and heterogeneous types of data and require better and more specific types of data visualization systems to help them. In this paper, we consider the data landscape that domain experts seek to understand, namely the set of datasets that are either currently available or could be obtained. Experts need to understand this landscape to triage which data analysis projects might be viable, out of the many possible research questions that they could pursue. We identify data reconnaissance and task wrangling as processes that experts undertake to discover and identify sources of data that could be valuable for some specific analysis goal. These processes have thus far not been formally named or defined by the research community. We provide formal definitions of data reconnaissance and task wrangling and describe how they relate to the data landscape that domain experts must uncover. We propose a conceptual framework with a four-phase cycle of acquire, view, assess, and pursue that occurs within three distinct chronological stages, which we call fog and friction, informed data ideation, and demarcation of final data. Collectively, these four phases embedded within three temporal stages delineate an expert’s progressively evolving understanding of the data landscape. We describe and provide concrete examples of these processes within the visualization community through an initial systematic analysis of previous design studies, identifying situations where there is evidence that they were at play. We also comment on the response of domain experts to this framework, and suggest design implications stemming from these processes to motivate future research directions. As technological changes will only keep adding unknown terrain to the data landscape, data reconnaissance and task wrangling are important processes that need to be more widely understood and supported by the data visualization tools. By articulating a concrete understanding of this challenge and its implications, our work impacts the design and evaluation of data visualization systems.","PeriodicalId":192801,"journal":{"name":"2019 IEEE Visualization Conference (VIS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114510196","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}