Vis -- Project: Suggestions

050065 VU Visualization - Project: suggestions

I'd like it if you come up with your own ideas! However, if you'd like to get some ideas, there are possibilities of projects in the areas of:

climate science (together with Niklas Röber, DKFZ, Hamburg)
The goal is the visualization of thermohaline circulation which is also known as the ocean conveyer belt. The task is finding it in a large data set and getting an idea what characteristics (speed, salt concentration, temperature, density) it might have in different parts of the ocean. The data set is a volumetric data set of size 800x400x50. You could use your own software to visualize it or some standard tools, like Avizo, ParaView, SimVis, Vapor, Visit. The key is finding the right transfer function!
summary of algorithmic performances (together with Martin Polascheck)
The lecture on Algorithm and Data Structures includes an assignment in which each student has to implement a data structure for sorting. At the end of the course, it is possible to compare one's own results with everybody else. However, there is a large set of graphs that are being produced and one looses the overview very quickly. The task is to create an interactive program to create a better overview of all the data as well as good interaction techniques to quickly drill down into the relevant details a user would like to see.
understanding image segmentations (together with Greg Mori, SFU, Canada)
We are trying to understand how image segmentation algorithms work by examining their results. The task is to create an interactive interface for the classification of images/scenes/objects based on a set of features. The interface should allow the user to choose the features for the recognition task, show the results, and evaluate its accuracy (the ground truth is given). The user should be able to combine different features, which should increase the accuracy. The question is -- how should different features be weighted? The data set and numerous other materials can be found at the SUN library page.
understanding tumor time series data (together with Katja Bühler, VRVis)
The data will be a number of 3D MRI images from a patient taken over the time span of months. The data will be aligned. The task is an understanding of how much and which parts of the brain have changed. Subtasks include a decision of visualizing 2D slices vs. 3D brainstructure, helping the user pick which timesteps should be compared, as well as ways on figuring out how the difference should be computed.
aiding in the modelling of engines (together with Harald Piringer, VRVis)
The data set consists of 1000 different simulations of an engine. Different design parameters (such as "EVO_shift", "IVO_shift", "ROI_shift", "Vane_p") have been changed for each simulation run. There are further parameters, such as "EnSpeed" and "Load_sig" that can take any values while the engine is running. The engineers are interested in minimizing ("TRAPPED-FUEL_MB", "COMBUSTION-NOISE_C of ENGINE 1") and maximizing (e.g. "TORQUE_PF") a number of other parameters. The simple question is -- what is the best parameter setting? Since there is not one single parameter setting that will fulfill all the constraints, can you build a system that let the engineer better understand the tradeoffs?
IEEE Vis + VAST contests
There are two different contests running right now. One is on the structure of a mouse brain and the other has three mini-challenges on movie openings, designs, network analysis. Further there is a BioVis re-Design challenge as well as a BioVis Data Contest. All of these have submission deadlines in July and August (the submission for the VU is still June 27) and I will pay the costs for attending the conference in October in Atlanta, USA (for one team member) for those teams that are able to finish a submission. I do believe these are doable for you!
still working on a project with autoPack
comparing dimensionality reduction techniques
comparing clustering techniques
... there may be others :)

Alternatively, you can pick a publicly available dataset (EuroStat, Open Data, etc.) and find an interesting story. Interesting data sets include, but are not limited to:

EuroStat -- data from the statistical office of the European Union.
Open Government Data (OCG) -- data for Vienna and Austria.

data.gov -- data from the US government.
Plenty of other sources around the webby web.

Visualization Contest

For a number of years, the Vis, InfoVis, and VAST conferences have created a visualization contest. For each contest a problem scenario together with the relevant data sets have been provided to the research community and a price has been awarded to the best visualization. Some of the problems have been quite challenging. However, for the most part, these are great problems to work on. Have a look:

Last modified: April 2013

Torsten Möller / torsten dot moeller AT univie DOT ac DOT at