Visualization Blog

Ideas, Papers and Thoughts on the field of Visualization

Archive for the ‘IEEE visualization 2007’ Category

Illustration-inspired visualization

with one comment

Illustration-inspired visualization has emerged as an interesting new way for visualization researchers to present data to practitioners and domain experts. Papers in this category are also categorized under “Illustrative visualization’, where the visualization is supposed to serve as a placeholder for an illustration of the underlying data. There have been some excellent papers by Bruckner et al., Li et al., Fischer et al. and
Correa et al. who have actually managed to generate impressive visualizations that look like illustrations.

Li et. \'07.

Last year at the IEEE Visualization 2007 conference, the ‘Best Application Paper’ award was given to a paper by Beyer et al. who discussed innovative ways of visualizing multimodal data for preoperative neurosurgical planning.
Illustration-inspired visualization also includes painterly visualization, which has been very effectively used for conveying multidimensional multivariate data. [There have been a couple of papers that explore illustrative information visualization, but links are not yet available and shall be posted as soon as they come online.]

Bruckner et al. \'05.

Lately, illustration-inspired visualization has been used for conveying change over time. Svakhine et al. used schlieren photography-inspired techniques and Joshi et al. used techniques inspired by comics to convey temporal context in a static visualization and have now conducted a user study to evaluate their techniques. Goldman et al. presented a storyboarding technique for video visualization at SIGGRAPH 2006. Chen et al. discussed a unique way of summarizing video content into visual signatures. Such summarizing techniques, techniques that provide temporal context are invaluable since the alternative is to keep looking at hours of video footage or look at 1000 snapshots of a time-varying dataset to observe changes.

There are many more such novel techniques that have allowed researchers to visualize application domain data in ways that their collaborators could not have imagined. But I wonder what is the reaction of the radiologists, neurosurgeons, hurricane experts, infovis users for whom the visualization techniques are intended in the first place? As visualization researchers and one who conducts research in the field of illustrative visualization, we generate a lot of wonderful techniques that allow for exploratory visualizations that reduce interaction and provide insight more intuitively. But, what is the end result of those
techniques? Are the domain experts happy and readily embracing those techniques for regular use? After working in two such data domains, I am of the opinion that illustrative visualization is a hard sell. Domain experts are usually used to rainbow colormaps (which we all know are unsuitable for visualizations) and have a preconceived idea of how their image should look. Showing them some new visualization technique seems
to end in “its interesting/beautiful, but we would like to have the standard visualization too.” There are always researchers who take the effort to conduct user studies to evaluate their illustrative techniques. I have previously written about my appreciation of researchers who go through the efforts of conducting a user study and I think it is quite essential. Whether the user studies are serving the purpose or not is probably a conversation for another day.

Illustrative visualization is a fascinating field that has endless avenues for visualizing data. Are we making a difference though? Any comments/suggestions or links to success stories would be greatly appreciated. I leave you with a fascinating capstone talk that I attended at IEEE Vis 2004 by none other than Pat Hanrahan. The talk is titled “Self illustrating Phenomena” and reaffirms my belief in the power of illustrative visualization.

Visualization panels and symposiums

leave a comment »

I’m sorry I never got around to finishing my review of the IEEE Visualization conference. It was a superb conference though. I would definitely encourage more activities like the

Meet the scientists panel

and the excellent

Impact of Data Visualization panel

I definitely think we need to make more people aware of the impact of visualization. We need more endeavors like the World Visualization Day that Robert Kosara has proposed. Ben Shneiderman and others are holding a Visualization Day at the City College of New York. The huge visualization group has organized a Visualization in the World symposium. [If anyone at UNCC reads this blog, please arrange to have a video of the wonderful talks that some of us cannot attend, thanks :) ]

The beauty of such efforts is that not only does it highlight the excellent accomplishments of our field, but also attracts the attention of domain experts who could use our findings. I think whats even more important is that through them new collaborations can be spawned that can bring about novel visualization challenges. The new avenues that such efforts may open are exciting to me.

If you know of any more such activities that are ongoing, please let me know.

Written by alark

April 25, 2008 at 9:20 am

IEEE Visualization 2007 conference review – part 1

with one comment

Hello all,

The IEEE Visualization 2007 conference concluded in Sacramento, CA on November 1, 2007. I had mentioned some of my favorite papers when the papers list was out. But, I found some more papers, panels and talks to be even more interesting than I had thought.

I guess its going to be a big post, since I’m going to go through all that I liked in the conference (and maybe some things that I didnt like).

  • The InfoVis keynote by Matt Ericson of New York Times was very well done. His group faces the daily challenge of communicating ideas using a visual representation of facts, stories and data. Unlike visualization researchers who assume that their audience will take the effort to understand their complex visualization, they at NYTimes cannot assume any such thing and have to develop ‘informational graphics’ for a reader. Fernanda Viégas from IBM Research has summarized some of the key points of the talk and also has managed to get access to some of his images. Here’s the link on infosthetics – http://infosthetics.com/archives/2007/10/infovis_keynote_matthew_ericson.html
  • The InfoVis session on Infovis for the Masses was one of my favorite sessions – The session chair was Ben Shneiderman who does such a fantastic job every time he talks. I attending his paper presentation last year and this year, as session chair, he setup the talks so well. The session contained a paper on Many Eyes (which as the readers of this blog may have observed, I really like :)) Fernanda Viégas, Martin Wattenberg and Frank van Hamm gave an excellent presentation. The second impressive talk of the session was on Scented Widgets by Wes Willett at UC Berkeley. Jock Mackinlay then showed a heavily demo-focussed presentation of the Show Me: Automatic Presentation for Visual Analysis paper. He was demoing the Tableau software from Tableau, which is the company he is now working for.
  • The Evaluation session was another fabulous session that had a great collection of superb papers. I liked each and every one of them.
  • Jeff Heer’s talk on Animated Transitions in Statistical Data Graphics discussed some work he did with George Robertson at Microsoft Research. He discussed some of the limitations of using animations that are extremely well documented by Barbara Tversky‘s paper – Animation: Can it facilitate? Visualizing any kind of change is extremely hard and he sort of weaved his way around the problems of animations and presented some great work. The user study was well conducted and I think its essential to be able to really stand up and defend your results in a quality conference such as InfoVis (pretty pictures obviously help) :) Some results obtained by them were not in favor of their techniques, but thats a result too and it was interesting to see that. In case you didnt already know, Jeff Heer is also the author of the popularly used Prefuse infovis toolkit.
  • This year was the second year they had the Art show at Vis and they showcased some really beautiful and insightful visualization.
  • This year the conference was also augmented by some excellent workshops on Knowledge-assisted visualization and VisSec – visualization for computer security. Smaller more focused workshops are an excellent venue for researchers to publish papers that might not get accepted due to the size of the idea or due to the fact that they’re too specialized for Infovis/Vis. The downside is that sometimes the ideas though great might not as polished and get accepted due to the low submission numbers at the workshop. I still give such workshops a big thumbs up and hope we see more of them.

There were some excellent tutorials at the conference this year. Some of the ones that I attended and my thoughts on the same

  • Visual Medicine tutorial – The morning session on Introduction to visual medicine though well presented is really something that does not need to be presented for the second part of the day on Advanced Visual Medicine to be understood. The advanced session was amazing and presented some phenomenal work that researchers were doing to use the power of medicine in improving doctors treat patients better in DT MRI, reconstructive surgery, perfusion data and multiresolution volume rendering.
  • Illustrative Display and Interaction in Visualization – This tutorial has become a mainstay of the conference and has included a great group of speakers over the years. The images produced present a compelling body of excellent work in the field of illustrative visualization. I have attended it in the past and have seen parts of it this time. The only minor complaint that I have is that the work is of limited use unless its being use in an application domain such as preoperative planning systems and other real world applications. In the European version of their tutorial at Eurographics 2006, they had a few other speakers who covered preoperative planning using such techniques.
  • GeoVisualization with Google Earth and GIS - This tutorial clearly conveyed that Google earth has a lot of wonderful applications than just flying around the world. The ability to add layers and interact with regions of interest was simply superb. Some datasets and files that they used during their tutorial can be downloaded here.

I realize now that I have many more thoughts that I’d like to add as regards the conference and so I shall add them in part 2. Coming soon!! Well you know how often I update this blog, so ….. soonish :)

IEEE Visualization 2007 papers posted

with one comment

Hello all,Sorry for not having written a blog entry in a while.

This year’s IEEE Visualization 2007 and the InfoVis 2007 papers have been announced. Some of you avid vis/infovis readers already knew that though, didnt you?

Similar to Tim Rowley’s SIGGRAPH papers link, , Yingcai Wu has been maintaining a site for IEEE Visualization papers that is updated fairly regularly.

The papers with updated links can be found at http://www.cse.ust.hk/~wuyc/Vis07.htm

The InfoVis 2007 papers link has been maintained by Hong Zhou and can be found at http://www.cse.ust.hk/~zhouhong/InfoVis07_paper.html

I will definitely be visiting the IEEE Visualization/InfoVis conference in October end and shall post my favorites then. But the papers that have already caught my attention are:

  1. Many Eyes: A Site for Visualization at Internet Scale
    Fernanda Viégas, Martin Wattenberg (Visual Communication Lab, IBM), Frank van Ham (TUE), Jesse Kriss, Matt McKeon (Visual Communication Lab, IBM)
  2. Illustrative Deformation for Data Exploration,
    Carlos Correa, Deborah Silver (The State University of New Jersey), Min Chen (Swansea University)
  3. Hotmap: Looking at Geographic Attention, Danyel Fisher (VIBE, Microsoft Research)
  4. Querying and Creating Visualizations by Analogy
    Carlos E. Scheidegger, Huy T. Vo, David Koop, Juliana Freire, Cláudio Silva
    (The University of Utah)

If you have any favorites of your own, feel free to add them to the comments. Shall read more papers as I get time and post my thoughts on the same.

Written by alark

September 12, 2007 at 8:59 am

Follow

Get every new post delivered to your Inbox.