VisLunch/Fall2008 - Revision history

Llins at 00:22, 20 November 2008

2008-11-20T00:22:59Z

← Older revision		Revision as of 00:22, 20 November 2008
Line 1:		Line 1:
	__NOTOC__		__NOTOC__

	== Coming up next Friday (10/03) ==		== Coming up next Friday (11/21) ==
	~~=== VIS 2008 papers: Session I =~~==

	~~Tomorrow~~ (~~Friday 10/03, 12pm~~) in ~~our VisLunch session <br~~/>		=== Towards Gigapixel displays : What?, How? and Why? (Chris Goodyer) ===
	~~we are going to see Carlos Scheidegger and David Koop <br/>~~
	~~presenting their VIS 2008 papers~~: ~~<br/>~~		Chris Goodyer from the University of Leeds in the UK will be
			talking about his recent work on multi-monitor displays:

			http://www.comp.leeds.ac.uk/ceg/IB3dviewer.htm

			Here is his web page:

	~~'''Revisiting Histograms and Isosurface Statistics''' <br/>~~		http://www.comp.leeds.ac.uk/ceg/
	~~Carlos Scheidegger, John Schreiner, Brian Duffy, Hamish Carr, Claudio Silva <br/>~~
	http://www.~~sci~~.~~utah~~.~~edu/~cscheid/vis2008~~/~~histograms <br~~/>

	~~''Abstract'' <br/>~~		Towards Gigapixel displays : What?, How? and Why?
	~~Recent results have shown a link between geometric properties of isosurfaces and statistical properties of the underlying sampled data. However, this has two defects~~: not all of the properties described converge to the same solution, and the statistics computed are not always invariant under isosurface-preserving transformations. We apply Federer's Coarea Formula from geometric measure theory to explain these discrepancies. We describe an improved substitute for histograms based on weighting with the inverse gradient magnitude, develop a statistical model that is invariant under isosurface-preserving transformations, and argue that this provides a consistent method for algorithm evaluation across multiple datasets based on histogram equalization. We use our corrected formulation to reevaluate recent results on average isosurface complexity, and ~~show evidence that noise is one cause of the discrepancy between the expected figure and the observed one.~~

	~~'''VisComplete: Automating Suggestions~~ for ~~Visualization Pipelines''' <br/>~~		Over the past couple of years there has been a growing trend for high
	~~David Koop~~, ~~Carlos E~~. ~~Scheidegger, Steven P~~. ~~Callahan~~, ~~Juliana Freire, Claudio T~~. ~~Silva <br/>~~		resolution display walls, intriguingly named "gigapixel displays" or
			"powerwalls". These are tiled multile projectors or multiple screens
			connected to a cluster. Besides looking visually impressive there are
			important research questions concerning what advantages there would be to
			using them over conventional large displays. In addition, the methods used
			for interacting with applications running across such large numbers of
			displays needs to be carefully considered.

	~~''Abstract'' <br/>~~		In this talk I shall summarize some of the work we have been doing at the
	~~Building visualization and analysis pipelines is a large hurdle in the adoption of visualization and workflow systems by domain scientists.~~ In this ~~paper, we propose techniques to help users construct pipelines by consensus—automatically suggesting completions based on a database~~ of ~~previously created pipelines. In particular,~~ we ~~compute correspondences between existing pipeline subgraphs from~~ the ~~database, and use these to predict sets~~ of ~~likely pipeline additions to a given partial pipeline~~. ~~By presenting these predictions in a carefully designed interface, users can create visualizations and other data products more efficiently because they can augment~~ their ~~normal work patterns with~~ the ~~suggested completions. We present an implementation~~ of ~~our technique in a publicly-available, open-source scientific workflow system~~ and ~~demonstrate efficiency gains in real-world situations.~~		University of Leeds on our 53-megapixel wall. I will discuss several of
			the software environments available for deploying applications along with
			their shortcomings. I shall also pose the important questions of "How big
			is big enough?" and conversely "How big is too big?"

	== Sessions ==		== Sessions ==
Line 206:		Line 216:
	VOL. 14, NO. 4, JULY/AUGUST 2008." <br/>		VOL. 14, NO. 4, JULY/AUGUST 2008." <br/>

	=== 10/03 - ~~Vis Conference~~: Carlos, ~~Dave ===~~		=== 10/03 VIS 2008 papers: Session I ===

			Tomorrow (Friday 10/03, 12pm) in our VisLunch session <br/>
			we are going to see Carlos Scheidegger and David Koop <br/>
			presenting their VIS 2008 papers: <br/>

			'''Revisiting Histograms and Isosurface Statistics''' <br/>
			Carlos Scheidegger, John Schreiner, Brian Duffy, Hamish Carr, Claudio Silva <br/>
			http://www.sci.utah.edu/~cscheid/vis2008/histograms <br/>

			''Abstract'' <br/>
			Recent results have shown a link between geometric properties of isosurfaces and statistical properties of the underlying sampled data. However, this has two defects: not all of the properties described converge to the same solution, and the statistics computed are not always invariant under isosurface-preserving transformations. We apply Federer's Coarea Formula from geometric measure theory to explain these discrepancies. We describe an improved substitute for histograms based on weighting with the inverse gradient magnitude, develop a statistical model that is invariant under isosurface-preserving transformations, and argue that this provides a consistent method for algorithm evaluation across multiple datasets based on histogram equalization. We use our corrected formulation to reevaluate recent results on average isosurface complexity, and show evidence that noise is one cause of the discrepancy between the expected figure and the observed one.

			'''VisComplete: Automating Suggestions for Visualization Pipelines''' <br/>
			David Koop, Carlos E. Scheidegger, Steven P. Callahan, Juliana Freire, Claudio T. Silva <br/>

			''Abstract'' <br/>
			Building visualization and analysis pipelines is a large hurdle in the adoption of visualization and workflow systems by domain scientists. In this paper, we propose techniques to help users construct pipelines by consensus—automatically suggesting completions based on a database of previously created pipelines. In particular, we compute correspondences between existing pipeline subgraphs from the database, and use these to predict sets of likely pipeline additions to a given partial pipeline. By presenting these predictions in a carefully designed interface, users can create visualizations and other data products more efficiently because they can augment their normal work patterns with the suggested completions. We present an implementation of our technique in a publicly-available, open-source scientific workflow system and demonstrate efficiency gains in real-world situations.

	=== 10/10 - Vis Conference: Jens, Geoff ===		=== 10/10 - Vis Conference: Jens, Geoff ===

Llins: /* 11/21 - Chris Goodyer */

2008-10-27T16:27:08Z

11/21 - Chris Goodyer

← Older revision		Revision as of 16:27, 27 October 2008
Line 216:		Line 216:
	=== 10/31 - Vis Highlights ===		=== 10/31 - Vis Highlights ===

	=== 11/21 - Chris Goodyer ===		=== 11/21 - Towards Gigapixel displays : What?, How? and Why? (Chris Goodyer) ===

	Chris Goodyer from the University of Leeds in the UK will be visiting on Nov 21.		Chris Goodyer from the University of Leeds in the UK will be visiting on Nov 21.

Llins at 16:26, 27 October 2008

2008-10-27T16:26:00Z

← Older revision		Revision as of 16:26, 27 October 2008
Line 227:		Line 227:
	http://www.comp.leeds.ac.uk/ceg/		http://www.comp.leeds.ac.uk/ceg/

			Towards Gigapixel displays : What?, How? and Why?

			Over the past couple of years there has been a growing trend for high
			resolution display walls, intriguingly named "gigapixel displays" or
			"powerwalls". These are tiled multile projectors or multiple screens
			connected to a cluster. Besides looking visually impressive there are
			important research questions concerning what advantages there would be to
			using them over conventional large displays. In addition, the methods used
			for interacting with applications running across such large numbers of
			displays needs to be carefully considered.

			In this talk I shall summarize some of the work we have been doing at the
			University of Leeds on our 53-megapixel wall. I will discuss several of
			the software environments available for deploying applications along with
			their shortcomings. I shall also pose the important questions of "How big
			is big enough?" and conversely "How big is too big?"

	== Mail List: vislunch@sci.utah.edu ==		== Mail List: vislunch@sci.utah.edu ==

Llins at 00:07, 9 October 2008

2008-10-09T00:07:25Z

← Older revision		Revision as of 00:07, 9 October 2008
Line 215:		Line 215:

	=== 10/31 - Vis Highlights ===		=== 10/31 - Vis Highlights ===

			=== 11/21 - Chris Goodyer ===

			Chris Goodyer from the University of Leeds in the UK will be visiting on Nov 21.
			He is talking about his recent work on multi-monitor displays:

			http://www.comp.leeds.ac.uk/ceg/IB3dviewer.htm

			Here is his web page:

			http://www.comp.leeds.ac.uk/ceg/


	== Mail List: vislunch@sci.utah.edu ==		== Mail List: vislunch@sci.utah.edu ==

Dakoop: /* 10/10 - Vis Conference: Jens Jeff */

2008-10-06T16:30:07Z

10/10 - Vis Conference: Jens Jeff

← Older revision		Revision as of 16:30, 6 October 2008
Line 208:		Line 208:
	=== 10/03 - Vis Conference: Carlos, Dave ===		=== 10/03 - Vis Conference: Carlos, Dave ===

	=== 10/10 - Vis Conference: Jens ~~Jeff~~ ===		=== 10/10 - Vis Conference: Jens, Geoff ===

	=== 10/17 - Fall Break (no lunch) ===		=== 10/17 - Fall Break (no lunch) ===

Llins: /* VIS 2008 papers Session I */

2008-10-02T21:52:04Z

VIS 2008 papers Session I

← Older revision		Revision as of 21:52, 2 October 2008
Line 2:		Line 2:

	== Coming up next Friday (10/03) ==		== Coming up next Friday (10/03) ==
	=== VIS 2008 papers Session I ===		=== VIS 2008 papers: Session I ===

	Tomorrow (Friday 10/03, 12pm) in our VisLunch session <br/>		Tomorrow (Friday 10/03, 12pm) in our VisLunch session <br/>

Llins: /* 09/26 - Valerio's suggested papers 2 */

2008-10-02T21:51:29Z

09/26 - Valerio's suggested papers 2

← Older revision		Revision as of 21:51, 2 October 2008
Line 186:		Line 186:
	''multigrid." <br/>		''multigrid." <br/>

	=== 09/26 - ~~Valerio's suggested papers 2~~ ===		=== 09/26 - Robust analysis of Vector Field Topology ===

			Tomorrow (Friday 09/26, 12pm) in our VisLunch session <br/>
			we are going to see Harsh Bhatia and Shreeraj Jadhav <br/>
			presenting: <br/>

			"Robust analysis of Vector Field Topology, <br/>

			We will be speaking about the combinatorial ways of analyzing topology <br/>
			of vector fields. More specifically, we will study the topology in terms of <br/>
			directed graphs called Morse Connection Graphs (MCG), & <br/>
			Entity Connection Graphs (ECG). We will do their comparative study, <br/>
			and discuss their robustness. We will also discuss efficient algorithms to <br/>
			generate the graphs. <br/>

			The algorithms are suggested by Guoning Chen, Konstantin Mischaikow, <br/>
			Robert S. Laramee, and Eugene Zhang in 'Efficient Morse Decompositions of Vector Fields', <br/>
			published at IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS, <br/>
			VOL. 14, NO. 4, JULY/AUGUST 2008." <br/>

	=== 10/03 - Vis Conference: Carlos, Dave ===		=== 10/03 - Vis Conference: Carlos, Dave ===

Llins: /* Poisson Editing: Past, Present, and Future */

2008-10-02T21:49:26Z

Poisson Editing: Past, Present, and Future

← Older revision		Revision as of 21:49, 2 October 2008
Line 2:		Line 2:

	== Coming up next Friday (10/03) ==		== Coming up next Friday (10/03) ==
	=== ~~Poisson Editing: Past, Present, and Future~~ ===		=== VIS 2008 papers Session I ===

	Tomorrow (Friday 10/03, 12pm) in our VisLunch session <br/>		Tomorrow (Friday 10/03, 12pm) in our VisLunch session <br/>

Llins at 21:48, 2 October 2008

2008-10-02T21:48:36Z

← Older revision		Revision as of 21:48, 2 October 2008
Line 1:		Line 1:
	__NOTOC__		__NOTOC__
	== Coming up next Friday (09/19) ==
			== Coming up next Friday (10/03) ==
	=== Poisson Editing: Past, Present, and Future ===		=== Poisson Editing: Past, Present, and Future ===
	~~Tomorrow (Friday 09/19, 12pm) in our VisLunch session <br/>~~
	~~we are going to see Michael Bellem and Brian Summa <br/>~~
	~~presenting: <br/>~~

	''~~"Poisson Editing: Past~~, ~~Present~~, ~~and Future~~''<br/>		Tomorrow (Friday 10/03, 12pm) in our VisLunch session <br/>
			we are going to see Carlos Scheidegger and David Koop <br/>
			presenting their VIS 2008 papers: <br/>

			'''Revisiting Histograms and Isosurface Statistics''' <br/>
			Carlos Scheidegger, John Schreiner, Brian Duffy, Hamish Carr, Claudio Silva <br/>
			http://www.sci.utah.edu/~cscheid/vis2008/histograms <br/>

			''Abstract'' <br/>
			Recent results have shown a link between geometric properties of isosurfaces and statistical properties of the underlying sampled data. However, this has two defects: not all of the properties described converge to the same solution, and the statistics computed are not always invariant under isosurface-preserving transformations. We apply Federer's Coarea Formula from geometric measure theory to explain these discrepancies. We describe an improved substitute for histograms based on weighting with the inverse gradient magnitude, develop a statistical model that is invariant under isosurface-preserving transformations, and argue that this provides a consistent method for algorithm evaluation across multiple datasets based on histogram equalization. We use our corrected formulation to reevaluate recent results on average isosurface complexity, and show evidence that noise is one cause of the discrepancy between the expected figure and the observed one.

	''~~Operating in the gradient-domain has become a standard <br/>~~		'''VisComplete: Automating Suggestions for Visualization Pipelines''' <br/>
	''~~method in image processing with applications in areas~~ <br/>		David Koop, Carlos E. Scheidegger, Steven P. Callahan, Juliana Freire, Claudio T. Silva <br/>
	~~''such as image compositing and tone-mapping~~. <br/>

	''~~Tomorrow we will cover the general Poisson method for~~ <br/>		''Abstract'' <br/>
	~~''editing images~~ in the ~~gradient~~ domain. ~~Also~~, we ~~will <br/>~~		Building visualization and analysis pipelines is a large hurdle in the adoption of visualization and workflow systems by domain scientists. In this paper, we propose techniques to help users construct pipelines by consensus—automatically suggesting completions based on a database of previously created pipelines. In particular, we compute correspondences between existing pipeline subgraphs from the database, and use these to predict sets of likely pipeline additions to a given partial pipeline. By presenting these predictions in a carefully designed interface, users can create visualizations and other data products more efficiently because they can augment their normal work patterns with the suggested completions. We present an implementation of our technique in a publicly-available, open-source scientific workflow system and demonstrate efficiency gains in real-world situations.
	~~''discuss~~ the ~~problems~~ with ~~this method when dealing <br/>~~
	~~''with gigapixel sized images~~. We ~~will look at how to <br/>~~
	~~''extend and accelerate this method using quadtrees~~, ~~<br/>~~
	''and ~~the most recent work~~ in ~~Poisson editing: streaming <br/>~~
	~~''multigrid~~.~~" <br/>~~

	== Sessions ==		== Sessions ==
Line 165:		Line 166:
	<br/>		<br/>

	=== 09/19 - ~~Valerio's suggested papers 1~~ ===		=== 09/19 - Poisson Editing: Past, Present, and Future ===

			Friday 09/19, 12pm, in our VisLunch session <br/>
			we are going to see Michael Bellem and Brian Summa <br/>
			presenting: <br/>

			''"Poisson Editing: Past, Present, and Future''<br/>

			''Operating in the gradient-domain has become a standard <br/>
			''method in image processing with applications in areas <br/>
			''such as image compositing and tone-mapping. <br/>

			''Tomorrow we will cover the general Poisson method for <br/>
			''editing images in the gradient domain. Also, we will <br/>
			''discuss the problems with this method when dealing <br/>
			''with gigapixel sized images. We will look at how to <br/>
			''extend and accelerate this method using quadtrees, <br/>
			''and the most recent work in Poisson editing: streaming <br/>
			''multigrid." <br/>

	=== 09/26 - Valerio's suggested papers 2 ===		=== 09/26 - Valerio's suggested papers 2 ===

Llins at 01:27, 19 September 2008

2008-09-19T01:27:03Z

← Older revision		Revision as of 01:27, 19 September 2008
Line 8:		Line 8:
	''"Poisson Editing: Past, Present, and Future''<br/>		''"Poisson Editing: Past, Present, and Future''<br/>

	Operating in the gradient-domain has become a standard <br/>		''Operating in the gradient-domain has become a standard <br/>
	method in image processing with applications in areas <br/>		''method in image processing with applications in areas <br/>
	such as image compositing and tone-mapping. <br/>		''such as image compositing and tone-mapping. <br/>

	Tomorrow we will cover the general Poisson method for <br/>		''Tomorrow we will cover the general Poisson method for <br/>
	editing images in the gradient domain. Also, we will <br/>		''editing images in the gradient domain. Also, we will <br/>
	discuss the problems with this method when dealing <br/>		''discuss the problems with this method when dealing <br/>
	with gigapixel sized images. We will look at how to <br/>		''with gigapixel sized images. We will look at how to <br/>
	extend and accelerate this method using quadtrees, <br/>		''extend and accelerate this method using quadtrees, <br/>
	and the most recent work in Poisson editing: streaming <br/>		''and the most recent work in Poisson editing: streaming <br/>
	multigrid." <br/>		''multigrid." <br/>

	== Sessions ==		== Sessions ==