Hsiang-Yun Wu, Aleksandr Amirkhanov, N. Grossmann
2021
Citations
0
Influential Citations
0
Citations
Journal
Vis. Informatics
Abstract
Building-up and running a university-based research group is a multi-faceted undertaking. The visualization working group at TU Wien (vis-group) has been internationally active over more than 25 years. The group has been acting in a competitive scientific setting where sometimes contradicting multiple objectives require trade-offs and optimizations. Research-wise the group has been performing basic and applied research in visualization and visual computing. Teaching-wise the group has been involved in undergraduate and graduate lecturing in (medical) visualization and computer graphics. To be scientifically competitive requires to constantly expose the group and its members to a strong international competition at the highest level. This necessitates to shield the members against the ensuing pressures and demands and provide (emotional) support and encouragement. Internally, the vis-group has developed a unique professional and social interaction culture: work and celebrate, hard and together. This has crystallized into a nested, recursive, and triangular organization model, which concretizes what it takes to make a research group successful. The key elements are the creative and competent vis-group members who collaboratively strive for (scientific) excellence in a socially enjoyable environment. © 2021 The Author(s). Published by Elsevier B.V. on behalf of ZhejiangUniversity and ZhejiangUniversity Press Co. Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 1. Overview and history The visualization working group (vis-group) started out with n interest in fractal geometry applied to computer-generated atural phenomena (e.g., landscapes, clouds). Fractal objects are n many cases closely related to dynamical systems, which we egan to visualize in about 1994. A small group of researchers nd (PhD) students formed the vis-group at the then Institute of omputer Graphics at TU Wien, Austria. For the next few years he visualization of dynamical systems and flow data was the ocus of our research. ∗ Corresponding author. E-mail addresses: hsiang.yun.wu@acm.org (H.-Y. Wu), amirkhanov@cg.tuwien.ac.at (A. Amirkhanov), ngrossmann@cg.tuwien.ac.at N. Grossmann), tklein@cg.tuwien.ac.at (T. Klein), dvdkouril@cg.tuwien.ac.at D. Kouřil), miao@cg.tuwien.ac.at (H. Miao), lluidolt@cg.tuwien.ac.at L.R. Luidolt), mindek@cg.tuwien.ac.at (P. Mindek), rraidou@cg.tuwien.ac.at R.G. Raidou), viola@cg.tuwien.ac.at (I. Viola), waldner@cg.tuwien.ac.at M. Waldner), groeller@cg.tuwien.ac.at (M.E. Gröller). ttps://doi.org/10.1016/j.visinf.2021.02.003 468-502X/© 2021 The Author(s). Published by Elsevier B.V. on behalf of Zhejiang Univ the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Flow visualization has been a major topic in scientific visualization with volume rendering being another intensely investigated area. In the late nineties we started a series of applied research projects in medical visualization together with a company partner (Tiani Medgraph later acquired by AGFA). We researched medical visualization techniques (e.g., maximum intensity projection, curved-planar reformation), which then were integrated into a company-developed medical workstation. These activities led to a long-lasting cooperation with radiologists from the general hospital (AKH) in Vienna in the area of CT-angiography. In 2000 we were one of the scientific proponents of the VRVis research center (VRVis, 2021b), which is now Austria’s leading research institution in the field of visual computing. With more than 70 employees, VRVis is engaged in innovative research in cooperation with industrial companies and universities. Over the years, the vis-group has been involved in many research projects at VRVis and a sizeable number of PhD students graduated based on these applied research activities. Our work in medical visualization brought us to several basic research projects and further cooperation projects with company partners like Philips Healthcare Eindhoven and General ersity and Zhejiang University Press Co. Ltd. This is an open access article under H.-Y. Wu, A. Amirkhanov, N. Grossmann et al. Visual Informatics 5 (2021) 76–84 B Fig. 1. Prenatal ultrasound diagnosis (https://www.youtube.com/watch?v= D7quHKgEuk). Fig. 2. Computer Graphics Forum (CGF) cover contest winner (Nanographics, 2021a). Electrics (Kretztechnik). With Philips we investigated colonoscopic and orthopedic magnetic resonance data for analysis, diagnosis, and evaluation. Between 2009 and 2015 purely company funded projects with General Electrics dealt with natural fetascopic rendering, i.e., high-quality rendering of noise-affected 4D ultrasound data. We investigated the robust and efficient applicability of advanced lighting and material effects for realtime display. The results were successfully commercialized by the company partner in their prenatal ultrasound diagnosis station (Fig. 1). Medical visualization led us to a basic investigation of volume processing and depiction. Volume rendering is characterized by a high sensitivity to small parameter variations and occlusion effects. We proposed several new, efficient, and highquality methods in illustrative visualization (Viola et al., 2006; Bruckner and Gröller, 2006) where we have been one of the leading groups in advancing this area. Tutorials on illustrative visualizations were organized at conferences like Eurographics and IEEE Visualization. The visualization results in the medical domain, brought us to other application areas where volumetric data are researched as well. This led to a longstanding cooperation, since 2004 until now, with the Research Group on Computed Tomography at the University of Applied Sciences Upper Austria, Wels. Threeand fourdimensional volumetric data are generated through X-ray industrial tomography for the non-destructive analysis of advanced materials like CFRP (carbon-fiber reinforced polymers) (Reh et al., 2013; Weissenböck et al., 2019). Due to the dual constellation of basic research in our group at TU Wien and applied research at VRVis, we have been contacted 77 Fig. 3. The current professorships at the vis-group. Fig. 4. The hybrid photo of the vis-group in 2020. Top row, from left to right, Dr. Haichao Miao, Prof. Ivan Viola, Prof. Renata Raidou, Prof. Manuela Waldner, M.Sc. David Kouřil. Bottom row, from left to right, Dr. Tobias Klein, M.Sc. Laura R. Luidolt, Dr. Peter Mindek, Prof. Meister Eduard Gröller, M.Sc. Aleksandr Amirkhanov, Dr. Hsiang-Yun Wu, M.Sc. Nicolas Grossmann. in 2005 by the University of Bergen (UiB), Norway. In a somewhat similar set-up they wanted to establish a visualization group at UiB as well. We have been involved in building up the visualization group there by participating in the selection process of the corresponding positions and giving lectures. More recently not only one data set, but a set or sequence of datasets is available in all areas of computational sciences, where the abundance of data makes the analysis and visualization especially challenging. The necessity of analyzing many datasets simultaneously, has resulted from several of our investigated application domains (non-destructive testing, medical analysis, visual analytics for decision support systems under uncertainty). We have abstracted the domain-specific context and proposed several visualization approaches in the area of comparative and cohort visualization (Ortner et al., 2017). The rapidly changing domain knowledge of biological nanostructures requires to include at least semi-automatic adjustment capabilities in the corresponding analysis tools (see Fig. 2). Also, the involved phenomena are often characterized by several types of complexities (massive multi-instance, dense, multi-scale), which requires new visual computing contributions to cope with these intricacies (Klein et al., 2018; Kouřil et al., 2019). The current vis-group comprises four professorships (Fig. 3) and researchers (Fig. 4 ), under a strongly connected ecosystem with students and researchers. The vis-group has had an average size of about 10 researchers over the years. About 50 Ph.D. students and 120 Master students graduated from the vis-group. A sizable number of vis-group students have already established successful research groups in the area of visual computing on their own. Five former Ph.D. students have founded start-up companies. The vis-group has been very internationally oriented with Ph.D. students from many different countries. We have tried and H.-Y. Wu, A. Amirkhanov, N. Grossmann et al. Visual Informatics 5 (2021) 76–84