Abstract Scientists working with uncertain data, such as climate simulations, medical images, or ensembles of physical simulations, regularly confront the problem of comparing observations, e.g., to identify similarities, differences, or patterns. Current approaches in comparative visualization of uncertain scalar fields mainly rely on juxtaposition of both data and uncertainties, where each is represented using, e.g., color mapping or volume rendering. While interpretation of uncertain scalar data from visual encodings is already cognitively challenging, comparison of uncertain fields without explicit visualization support adds a further layer of complexity. In this paper, we present a theoretical framework to devise and describe a class of techniques that directly visualize differences between two or more uncertain scalar fields in a single image. We model each such technique as a combination of one or more interpolation stages, with the application of distance measures on random variables to the resulting distributions, and an appropriate visual encoding. Our framework captures existing methods and lends itself well to formulating new comparative visualization techniques for uncertain data for different visualization scenarios. Furthermore, by modeling uncertain scalar field differences as random variables themselves, we enable additional opportunities for comparison. We demonstrate the usefulness of our framework and its properties by applying it to effective comparative visualization techniques for several synthetic and real-world data sets.

When learning a motor skill it is helpful to get corrective feedback from an instructor. This will support the learner to execute the movement correctly. With modern technology, it is possible to provide this feedback via mixed reality. In most cases, this involves visual cues to help the user understand the corrective feedback. We analyzed recent research approaches utilizing visual cues for feedback in mixed reality. The scope of this article is visual feedback for motor skill learning, which involves physical therapy, exercise, rehabilitation etc. While some of the surveyed literature discusses therapeutic effects of the training, this article focuses on visualization techniques. We categorized the literature from a visualization standpoint, including visual cues, technology and characteristics of the feedback. This provided insights into how visual feedback in mixed reality is applied in the literature and how different aspects of the feedback are related. The insights obtained can help to better adjust future feedback systems to the target group and their needs. This article also provides a deeper understanding of the characteristics of the visual cues in general and promotes future, more detailed research on this topic.

Abstract In this work, we propose a visual analytics system to analyze deep reinforcement learning (deepRL) models working on the track reconstruction of charged particles in the field of particle physics. The data of these charged particles are in the form of point clouds with high-dimensional features. We use one of the existing post hoc saliency methods of explainable artificial intelligence (XAI) and extend its adaptation to compute saliency attributions for the input data corresponding to the output of the model. Our proposed system helps users to explore these saliency attributions corresponding to the high-dimensional input data of the machine learning model and interpret the decision-making process of the model. In particular, we provide the users with multiple task-oriented components, different types of linked views and interactive tools to analyze the model. We explain how to use the system by outlining a typical user workflow and demonstrate the system’s usefulness using several case studies which address specific analysis tasks.

We discuss the concept of directness in the context of spatial interaction with visualization. In particular, we propose a model that allows practitioners to analyze and describe the spatial directness of interaction techniques, ultimately to be able to better understand interaction issues that may affect usability. To reach these goals, we distinguish between different types of directness. Each type of directness depends on a particular mapping between different spaces, for which we consider the data space, the visualization space, the output space, the user space, the manipulation space, and the interaction space. In addition to the introduction of the model itself, we also show how to apply it to several real-world interaction scenarios in visualization, and thus discuss the resulting types of spatial directness, without recommending either more direct or more indirect interaction techniques. In particular, we will demonstrate descriptive and evaluative usage of the proposed model, and also briefly discuss its generative usage.

Linear approximation plays an important role in many areas employing numerical algorithms. Particularly in the field of vector field visualization, it is the basis of widely used techniques. In this paper, we introduce two methods to extract areas in two- and three-dimensional vector fields that are connected to linear flow behavior. We propose a region-growing algorithm that extracts the linear neighborhood for a certain position. The region is characterized by linear flow behavior up to a user-defined approximation threshold. While this first method computes the size of a region given the mentioned threshold, our second method computes the quality of a linear approximation given a user-defined n-ring neighborhood. The scalar field resulting from the second method is, therefore, called affine linear approximation error. Isosurfaces of this field show regions of close-to-linear and non-linear flow behavior. We demonstrate the expressiveness and discuss the properties of the extracted regions using analytical examples and several datasets from the domain of computational fluid dynamics (CFD).

Scientists, engineers and physicians are used to analyze 3D data with slice-based visualizations. Radiologists for example are trained to read slices of medical imaging data. Despite the numerous examples of sophisticated 3D rendering techniques, domain experts, who still prefer slice-based visualization do not consider these to be very useful. Since 3D renderings have the advantage of providing an overview at a glance, while 2D depictions better serve detailed analyses, it is of general interest to better combine these methods. Recently there have been attempts to bridge this gap between 2D and 3D renderings. These attempts include specialized techniques for volume picking in medical imaging data that result in repositioning slices. In this paper, we present a new volume picking technique called WYSIWYP ("what you see is what you pick") that, in contrast to previous work, does not require pre-segmented data or metadata and thus is more generally applicable. The positions picked by our method are solely based on the data itself, the transfer function, and the way the volumetric rendering is perceived by the user. To demonstrate the utility of the proposed method, we apply it to automated positioning of slices in volumetric scalar fields from various application areas. Finally, we present results of a user study in which 3D locations selected by users are compared to those resulting from WYSIWYP. The user study confirms our claim that the resulting positions correlate well with those perceived by the user.

Abstract We present an algorithm that allows stream surfaces to recognize and adapt to vector field topology. Standard stream surface algorithms either refine the surface uncontrolled near critical points which slows down the computation considerably and may lead to a poor surface approximation. Alternatively, the concerned region is omitted from the stream surface by severing it into two parts thus generating an incomplete stream surface. Our algorithm utilizes topological information to provide a fast, accurate, and complete triangulation of the stream surface near critical points. The required topological information is calculated in a preprocessing step. We compare our algorithm against the standard approach both visually and in performance.

Stream surfaces are an intuitive approach to represent 3D vector fields. In many cases, however, they are challenging objects to visualize and to understand, due to a high degree of self-occlusion. Despite the need for adequate rendering methods, little work has been done so far in this important research area. In this paper, we present an illustrative rendering strategy for stream surfaces. In our approach, we apply various rendering techniques, which are inspired by the traditional flow illustrations drawn by Dallmann and Abraham \& Shaw in the early 1980s. Among these techniques are contour lines and halftoning to show the overall surface shape. Flow direction as well as singularities on the stream surface are depicted by illustrative surface streamlines. ;To go beyond reproducing static text book images, we provide several interaction features, such as movable cuts and slabs allowing an interactive exploration of the flow and insights into subjacent structures, e.g., the inner windings of vortex breakdown bubbles. These methods take only the parameterized stream surface as input, require no further preprocessing, and can be freely combined by the user. We explain the design, GPU-implementation, and combination of the different illustrative rendering and interaction methods and demonstrate the potential of our approach by applying it to stream surfaces from various flow simulations. ;

Abstract We introduce a novel technique for the construction of smooth stream surfaces of 4th order precision. While common stream surface techniques use linear interpolation for generating seed points for new streamlines in the refinement phase, we use Hermite interpolation. The derivatives needed for Hermite interpolation are obtained by integration along the streamlines. This yields stream surfaces of4th order precision. Additionally, we analyse the accuracy ofthe well known Hultquist approach and our new algorithm and proof that Hultquist's method is exact for linear vector fields. We compare both methods using the well known distance based and a novel error based refinement strategy. Our resulting surface is C1‐continuous, enabling improved rendering among other benefits.

Abstract The notions of Finite‐Time Lyapunov Exponent (FTLE) and Lagrangian Coherent Structures provide a strong framework for the analysis and visualization of complex technical flows. Their definition is simple and intuitive, and they are built on a deep theoretical foundation. We apply these concepts to enable the analysis of flows in the immediate vicinity of the boundaries of flow‐embedded objects by limiting the Lagrangian analysis to surfaces closely neighboring these boundaries. To this purpose, we present an approach to approximate FTLE fields over such surfaces. Furthermore, we achieve an effective depiction of boundary‐related flow structures such as separation and attachment over object boundaries and specific insight into the surrounding flow using several specifically chosen visualization techniques. We document the applicability of our methods by presenting a number of application examples.

Understanding fluid flow data, especially vortices, is still a challenging task. Sophisticated visualization tools help to gain insight. In this paper, we present a novel approach for the interactive comparison of scalar fields using isosurfaces, and its application to fluid flow datasets. Features in two scalar fields are defined by largest contour segmentation after topological simplification. These features are matched using a volumetric similarity measure based on spatial overlap of individual features. The relationships defined by this similarity measure are ranked and presented in a thumbnail gallery of feature pairs and a graph representation showing all relationships between individual contours. Additionally, linked views of the contour trees are provided to ease navigation. The main render view shows the selected features overlapping each other. Thus, by displaying individual features and their relationships in a structured fashion, we enable exploratory visualization of correlations between similar structures in two scalar fields. We demonstrate the utility of our approach by applying it to a number of complex fluid flow datasets, where the emphasis is put on the comparison of vortex related scalar quantities.

Modern unsteady (multi-)field visualizations require an effective reduction of the data to be displayed. From a huge amount of information the most informative parts have to be extracted. Instead of the fuzzy application dependent notion of feature, a new approach based on information theoretic concepts is introduced in this paper to detect important regions. This is accomplished by extending the concept of local statistical complexity from finite state cellular automata to discretized (multi-)fields. Thus, informative parts of the data can be highlighted in an application-independent, purely mathematical sense. The new measure can be applied to unsteady multifields on regular grids in any application domain. The ability to detect and visualize important parts is demonstrated using diffusion, flow, and weather simulations.

We present, extend, and apply a method to extract the contribution of a subregion of a data set to the global flow. To isolate this contribution, we decompose the flow in the subregion into a potential flow that is induced by the original flow on the boundary and a localized flow. The localized flow is obtained by subtracting the potential flow from the original flow. Since the potential flow is free of both divergence and rotation, the localized flow retains the original features and captures the region-specific flow that contains the local contribution of the considered subdomain to the global flow. In the remainder of the paper, we describe an implementation on unstructured grids in both two and three dimensions for steady and unsteady flow fields. We discuss the application of some widely used feature extraction methods on the localized flow and describe applications like reverse-flow detection using the potential flow. Finally, we show that our algorithm is robust and scalable by applying it to various flow data sets and giving performance figures.

We present a method to extract and visualize vortices that originate from bounding walls of three-dimensional time-dependent flows. These vortices can be detected using their footprint on the boundary, which consists of critical points in the wall shear stress vector field. In order to follow these critical points and detect their transformations, affected regions of the surface are parameterized. Thus, an existing singularity tracking algorithm devised for planar settings can be applied. The trajectories of the singularities are used as a basis for seeding particles. This leads to a new type of streak line visualization, in which particles are released from a moving source. These generalized streak lines visualize the particles that are ejected from the wall. We demonstrate the usefulness of our method on several transient fluid flow datasets from computational fluid dynamics simulations.

In this paper we present an approach to the analysis of the contribution of a small subregion in a dataset to the global flow. To this purpose, we subtract the potential flow that is induced by the boundary of the sub-domain from the original flow. Since the potential flow is free of both divergence and rotation, the localized flow field retains the original features. In contrast to similar approaches, by making explicit use of the boundary flow of the subregion, we manage to isolate the region-specific flow that contains exactly the local contribution of the considered subdomain to the global flow. In the remainder of the paper, we describe an implementation on unstructured grids in both two and three dimensions. We discuss the application of several widely used feature extraction methods on the localized flow, with an emphasis on topological schemes.

It is a challenging task to visualize the behavior of time-dependent 3D vector fields. Most of the time an overview of unsteady fields is provided via animations, but, unfortunately, animations provide only transient impressions of momentary flow. In this paper we present two approaches to visualize time varying fields with fixed geometry. Path lines and streak lines represent such a steady visualization of unsteady vector fields, but because of occlusion and visual clutter it is useless to draw them all over the spatial domain. A selection is needed. We show how bundles of streak lines and path lines, running at different times through one point in space, like through an eyelet, yield an insightful visualization of flow structure ("eyelet lines"). To provide a more intuitive and appealing visualization we also explain how to construct a surface from these lines. As second approach, we use a simple measurement of local changes of a field over time to determine regions with strong changes. We visualize these regions with isosurfaces to give an overview of the activity in the dataset. Finally we use the regions as a guide for placing eyelets.