Visualization Multi-Pipeline for Communicating Biology

Information

• Publication Type: Journal Paper with Conference Talk
• Workgroup(s)/Project(s):
  • Visual Computing: Illustrative Visualization
• Date: 2017
• Journal: IEEE Transactions on Visualization and Computer Graphics
• Volume: 24
• Number: 1
• Location: Phoenix, Arizona
• Lecturer: Peter Mindek
• Event: IEEE Vis 2017
• Conference date: 29. September 2017 – 7. October 2017
• Keywords: Biological visualization, remote rendering, public dissemination

Abstract

We propose a system to facilitate biology communication by developing a pipeline to support the instructional visualization of heterogeneous biological data on heterogeneous user-devices. Discoveries and concepts in biology are typically summarized with illustrations assembled manually from the interpretation and application of heterogenous data. The creation of such illustrations is time consuming, which makes it incompatible with frequent updates to the measured data as new discoveries are made. Illustrations are typically non-interactive, and when an illustration is updated, it still has to reach the user. Our system is designed to overcome these three obstacles. It supports the integration of heterogeneous datasets, reflecting the knowledge that is gained from different data sources in biology. After pre-processing the datasets, the system transforms them into visual representations as inspired by scientific illustrations. As opposed to traditional scientific illustration these representations are generated in real-time - they are interactive. The code generating the visualizations can be embedded in various software environments. To demonstrate this, we implemented both a desktop application and a remote-rendering server in which the pipeline is embedded. The remote-rendering server supports multi-threaded rendering and it is able to handle multiple users simultaneously. This scalability to different hardware environments, including multi-GPU setups, makes our system useful for efficient public dissemination of biological discoveries.

Additional Files and Images

Weblinks

• VIS presentation

BibTeX

@article{mindek-2017-marion,
  title =      "Visualization Multi-Pipeline for Communicating Biology",
  author =     "Peter Mindek and David Kou\v{r}il and Johannes Sorger and
               David Toloudis and Blair Lyons and Graham Johnson and Eduard
               Gr\"{o}ller and Ivan Viola",
  year =       "2017",
  abstract =   "We propose a system to facilitate biology communication by
               developing a pipeline to support the instructional
               visualization of heterogeneous biological data on
               heterogeneous user-devices. Discoveries and concepts in
               biology are typically summarized with illustrations
               assembled manually from the interpretation and application
               of heterogenous data. The creation of such illustrations is
               time consuming, which makes it incompatible with frequent
               updates to the measured data as new discoveries are made.
               Illustrations are typically non-interactive, and when an
               illustration is updated, it still has to reach the user. Our
               system is designed to overcome these three obstacles. It
               supports the integration of heterogeneous datasets,
               reflecting the knowledge that is gained from different data
               sources in biology. After pre-processing the datasets, the
               system transforms them into visual representations as
               inspired by scientific illustrations. As opposed to
               traditional scientific illustration these representations
               are generated in real-time - they are interactive. The code
               generating the visualizations can be embedded in various
               software environments. To demonstrate this, we implemented
               both a desktop application and a remote-rendering server in
               which the pipeline is embedded. The remote-rendering server
               supports multi-threaded rendering and it is able to handle
               multiple users simultaneously. This scalability to different
               hardware environments, including multi-GPU setups, makes our
               system useful for efficient public dissemination of
               biological discoveries. ",
  journal =    "IEEE Transactions on Visualization and Computer Graphics",
  volume =     "24",
  number =     "1",
  keywords =   "Biological visualization, remote rendering, public
               dissemination",
  URL =        "https://www.cg.tuwien.ac.at/research/publications/2017/mindek-2017-marion/",
}