VizWorld.com

Indirect Multi-Touch Interaction for Brushing in Parallel Coordinates from Robert Kosara on Vimeo.

Robert Kosara was inspired by the clever multi-touch capabilities of his new Macbook pro, and found a great use for it when dealing with selections in Parallel Coordinates graphs.  While writing a paper for an upcoming conference about it, he was unable to include graphics so instead linked to a video he squirreled away on Vimeo.  Now that video is getting some attention (before the paper was printed), and he’s taken the opportunity to talk about it on his site.

If you’ve used a MacBook or MacBook Pro, you know how simple and intuitive the two-finger scrolling is. Similarly, there are some interactions on the iPhone that are so obvious (like the pinch for zooming) that it’s tough to remember how we did things before touch screens. I wanted to create a similar experience for brushing in parallel coordinates.

He also released a tiny demo app to show how it works, available at his site to try.

via Multi-touch Brushing for Parallel Coordinates | eagereyes.

Science algorithm, paper

Andy Woodruff, Robert Roth, and Zachary Johnson have a new paper in the latest issue of The Cartography Journal of the British Cartographic Society called “Value-by-alpha Maps: An Alternative Technique to the Cartogram”.

Value-by-alpha maps (hereafter shortened to VBA), like everything noble and good, have their roots in somebody complaining about something on the internet—me, about election cartograms. Seeking an alternative to what I think are ugly and unreadable election results cartograms, I worked with my Axis Maps dudes to create a 2008 U.S. election map that used transparency rather than size to vary the visual impact of map units, thinking that avoiding the distortion of these units into unrecognizable sizes and shapes would make the map easier to read.

One of the more interesting contributions in the paper is their solution to “Johnson’s Cube” where they map the 3 axes to Visual Equalisation, Topology Preservation, and Shape Preservation and show cartographic rendering techniques to illustrate the various combinations.

via Value-by-alpha maps | Cartogrammar.

Science algorithm, cartography, paper, research

A new scientific paper by Danny Holten and Jarke J. van Wijk entitled “A User Study on Visualizing Directed Edges in Graphs” (Download, View), looks at several algorithms for representing groups of directional lines to try and find the least confusing methods for users.

Their experiments consisted of testing the different visual techniques (or combinations of the techniques), on which participants performed specific tasks in which they had to answer whether or not there were directed connection from one point to another in a randomly generated graph. Response times and accuracy were measured and analyzed. The different techniques tested were: “arrow”, “light-to-dark”, “dark-to-light”, “green-to-red”, “curved”, and “tapered”.

via What is the Best Way to Represent Directionality in Network Visualizations? – information aesthetics.

Science paper

A research paper from IBM analyzes a FLOP-Intensive algorithm (dot-products and additions of 2D matrices of single-precision floating point values), and finds that the CPU can actually beat the GPU versions.  I’ll let the abstract do the talking (although I’ve reformatted it for readability).

We implement this algorithm on a nVidia GTX 285 GPU using CUDA, and also parallelize it for the Intel Xeon (Nehalem) and IBM Power7 processors, using both manual and automatic techniques. Pthreads and OpenMP with SSE and VSX vector intrinsics are used for the manually parallelized version, while a state-of-the-art optimization framework based on the polyhedral model is used for automatic compiler parallelization and optimization.

The performance of this algorithm on the nVidia GPU suffers from:

1. a smaller shared memory,
2. unaligned device memory access patterns,
3. expensive atomic operations, and
4. weaker single-thread performance.

On commodity multi-core processors, the application dataset is small enough to fit in caches, and when parallelized using a combination of task and short-vector data parallelism (via SSE/VSX) or through fully automatic optimization from the compiler, the application matches or beats the performance of the GPU version.

The primary reasons for better multi-core performance include larger and faster caches, higher clock frequency, higher on-chip memory bandwidth, and better compiler optimization and support for parallelization. The best performing versions on the Power7, Nehalem, and GTX 285 run in 1.02s, 1.82s, and 1.75s, respectively. These results conclusively demonstrate that, under certain conditions, it is possible for a FLOP-intensive structured application running on a multi-core processor to match or even beat the performance of an equivalent GPU version.

Not really surprised to see the Power7 perform the best (Given that it’s IBM’s chip and IBM engineers at the helm here).  If you read the paper, you’ll see that they use a 500×500 image (4MB in size) with matrices that require 250KB of space (page 6).  This won’t fit into the cache of the GTX285 so they spend much time paging data in and out.

I’ld be very curious to see if the new Fermi GTX480 changes this any.

via IBM Research | Technical Paper Search | Believe it or Not! Multicore CPUs can Match GPUs for FLOP-intensive Applications!|(Search Reports).

Science algorithm, ibm, paper

Many people thing computer graphics are a recent thing, but the math behind recreating reality in a virtual world is in fact long been a fascination with scholars.  Real-Time Rendering has built a list of some papers dating back over 300 years from such names as Isaac Newton and James Maxwell that built the foundation of modern rendering technologies.

The Royal Society (full name: Royal Society of London for the Improvement of Natural Knowledge) is marking the start of its 350th year by putting pdf versions of 60 notable papers from its journal, Philosophical Transactions (founded in 1665) on the web. Although all the selected papers are crucial to the history of science, I wanted to call out those particularly related to the fundamentals of rendering.

via Real-Time Rendering · Rendering-related papers of the past 350 years.

Science history, paper, reference

The ACM has published the list of accepted papers at the upcoming SIGGRAPH Asia conference, several of which are available in pre-print PDF form.

Leading international experts from Asia and beyond present peer-reviewed research in physical simulation, animation control, real-time and photo-realistic rendering, geometric and urban modeling, hair capture and styling, texturing, image and video processing and resizing, GPU algorithms, and sound. One session of four papers includes duplicate presentations in Japanese.

via Technical Papers > For Attendees > SIGGRAPH Asia 2009.

Graphics, Science conference, paper, siggraph

An interesting paper presented here at VisWeek, as part of VAST, comes from Christopher Collins Research.  They parsed a database of 600,000 US Circuit Court decisions over 50 years and created a text-visualization tool they call “Parallel Tag Clouds”.

The visualization technique combines graphical elements from parallel coordinates and traditional tag clouds to provide rich overviews of a document collection while acting as an entry point for exploration of individual texts. We augment basic parallel tag clouds with a details-in-context display and an option to visualize changes over a second facet of the data, such as time. We also address text mining challenges such as selecting the best words to visualize, and how to do so in reasonable time periods to maintain interactivity.

The full paper is available at their site.  See their demonstration video after the break.

via Christopher Collins: Research.

Read more…

Science paper, research, tag clouds, visweek

Marian Dörk has placed a pair of PDF’s online from his presentation yesterday at the “Vis on the Web” discussion where he discusses the failures of Ranked Snippets (like Google Search Results) and Thumbnails.

As the Web continues to grow immensely, we think a logical next step is developing more visual and interactive ways for exploring the Web. However, most visual search interfaces that are currently appearing do not make use of information visualization as a way to aggregate or summarize information. Instead we see playful graphical interfaces or previews of visual Web resources that employ ranking algorithms to reduce the number of displayed items.

His paper (1 page PDF) and Presentation Slides (PDF) are available online.

Science paper, visweek

Robin Strand & Stefan Gustavson have published some source-code, binaries, and documentation on a new Distance Transform algorithm that works with anti-aliased grayscale images (similar to OpenGL’s existing depthmap) rather than the super-high resolution images required by other methods.  Their abstract:

We present a modified distance transform measure for use with anti-aliased grayscale images of arbitrary contours. The modified measure can be used in any vector-propagation Euclidean distance transform, and provides a much improved distance accuracy compared to a traditional binary image transform. We also modify the fast-marching method to get a smoother distance transform.

You can visit their modest website to visit the forums and read the pre-print version of the paper as a PDF.

Science algorithm, paper

A new paper from the Finnish Geodesic Institute and the Helsinki Institute of Technology discusses a new method for computed 3D models of an environment using photogrammetry and laser scanning but adds in compensation for orientation problems.

Comprehensive 3D modeling of our environment requires integration of terrestrial and airborne data, which is collected, preferably, using laser scanning and photogrammetric methods. However, integration of these multi-source data requires accurate relative orientations. In this article, two methods for solving relative orientation problems are presented.

The entire paper is available as a PDF.

via Abstract: Orientation of Airborne Laser Scanning Point Clouds with Multi-View, Multi-Scale Image Blocks – MDPI.

Science paper, photogrammetry