Talk:Computer graphics/Archive 2

Mission

I'm putting back the following note on the "mission" for this page. I feel very strongly that this is the right thing to do and that there needs to be a page of this type on the Wikipedia - please do not merge it again without further discussion. Trevorgoodchild 05:49, 21 March 2007 (UTC)[reply]

Guidelines for this article

This article has recently experienced a "reboot," to focus on the acadmic discipline of Computer Graphics rather than applications of computer graphics to movies and video games (as these topics are covered extensively by other Wikipedia pages). It has been modeled after the general article on physics -- the idea is to give a broad overview of the problem, the connection of computer graphics to other disciplines, and some history of the field as a whole. You may want to take a look at the physics article before editing this one to get an idea of what a good overview article looks like. Some sections are left sparse or empty as placeholders for future information - please try to fill these in if you can. I have tried to maintain a broad perspective while linking to specialized pages when appropriate.

Specific information about graphics sub-topics is abundant on the Wikipedia and does not need to be repeated here!

Additionally, if you'd like to write something about, e.g., graphics specific to the movie industry or graphics specific to the video game industry, I would encourage you to contribute to the CGI page, one of the video games subpages, or to create a specialized page.

I've also weeded out links which are of microscopic relevance (e.g., links to the Macromedia Flash SDK) - let's try to keep it this way. Trevorgoodchild 05:49, 21 March 2007 (UTC)[reply]

This article needs massive attention before it will be any semblance of the kind of article that physics is. The information needs to be referenced, not one editor's personal experience/predilection. Most of the information here was coverred in 3D computer graphics before the merge, and the information here is not particularly informative outside of that context. The other article is much more "general" than this one is, and has a much larger editor base. Adam McCormick 06:10, 21 March 2007 (UTC)[reply]

Correct, this article does need a lot of work. But merging it with a more generic article is not the way to get there. The article on 3D graphic design/3D computer graphics does not make the distinction between the people who create graphics technology and those who use it. Merging the Computer Graphics page with the 3D graphic design page would be like merging the Physics page with some page on engineering. One is the science, the other is the application. Trevorgoodchild 06:26, 21 March 2007 (UTC)[reply]

Thanks to everyone (esp. Alanbly) for keeping this page in good shape, removing irrelevant links, etc. It's a tough page to organize since if you don't know a lot about CG it's easy to confuse theory with applications, but you guys are doing a great job! Trevorgoodchild 14:21, 4 October 2007 (UTC)[reply]

Boundary representations - citation needed

Is there anyone out there who doesn't agree that boundary representations (i.e., surface meshes) are the most common object representation in graphics? —The preceding unsigned comment was added by Trevorgoodchild (talk • contribs) 06:30, 21 March 2007 (UTC).[reply]

Those who know nothing about CG don't know that it is. They also don't know what a surface mesh is. ALL info on wikipedia is supposed to be attributable. Adam McCormick 06:38, 21 March 2007 (UTC)[reply]

Ok, but what do you suggest I do given that A. we both know it's true and B. there's not really any great source to cite? Trevorgoodchild 06:41, 21 March 2007 (UTC)[reply]

P.S. Just because all info on Wikipedia should be citable doesn't mean all info should be cited. Can you imagine what it would be like if everything was cited? It would be a mess! Trevorgoodchild 06:42, 21 March 2007 (UTC)[reply]

Also, the argument that "those who know nothing about CG don't know that it is" is not a valid one. Have you been to many of the math pages? There are plenty of things that are opaque to those not familiar with the area the article addresses. Not every page can start from scratch. Trevorgoodchild 06:49, 21 March 2007 (UTC)[reply]

I just suggest adding a sentence or two describing what "boundary representations" and "surface meshes" are before declaring them the most common object representation. Adam McCormick 16:26, 21 March 2007 (UTC)[reply]

The words "boundary representation" and "polygonal mesh" both link to their respective wiki pages, which I thought was the standard thing to do on Wikipedia...Trevorgoodchild 16:56, 21 March 2007 (UTC)[reply]

Possible Addition

All, I have a paper I wrote on CG and it includes a section on viewing transforms that may suffice it's here any thoughts on grafting some of it into the article? Adam McCormick 01:01, 3 March 2007 (UTC)[reply]

I don't think it's appropriate to take anything from or mention any specific publications on this page. After all, there are many thousand peer-reviewed computer graphics publicationsJhhays 20:24, 17 April 2007 (UTC)[reply]

Yes but those are not public domain (so they cannot be directly drawn from) where this is (as I have consented to adding anything from it). This discussion may belong more to 3d computer graphics than this article though. Adam McCormick 17:22, 19 April 2007 (UTC)[reply]

Major Subfields

I edited the major subfields to include "Imaging". This follows the arrangement of the SIGGRPAH 2007 papers committee which includes area coordinates for Geometry, Animation, Rendering, and Imaging. (see http://www.siggraph.org/s2007/presenters/papers/review.html ) A large amount of SIGGRAPH papers do not concern Geometry, animation, or rendering. This "Imaging" category includes the large number of computer graphics research projects which focus on image acquisition, image editing, or image processing. Jhhays 19:44, 17 April 2007 (UTC)jhhays[reply]

Added researchers

I added some notable researchers. I see that the previous list was based off of http://graphics.stanford.edu/~fedkiw/, which seems somewhat reasonable, but that page didn't include Shree Nayar (who had 6 siggraph papers last year, and has a G number of 5). The wiki page is also inconsistent about which researchers from that list have been included. Some people with G numbers of 2 are included and some are not. I also added Jessica Hodgins to the wiki page, she has served as the papers chair for SIGGRAPH and has a G number of 2. Jhhays 20:05, 17 April 2007 (UTC)[reply]

I'm also removing some researchers that don't seem to have any computer graphics related publications, for instance "Michael Crandall" and a few others that seem to focus on applied math or computational physics. But by all means tell me if I went to far.Jhhays 20:11, 17 April 2007 (UTC)[reply]

I think this list as it stands is somewhat lengthy and could probably stand to be trimmed extensively. At very least it needs to be heavily reorganized. This should be a list of the founders of CG research and the very notable. There needs to be a higher standard set here. Adam McCormick 00:35, 24 April 2007 (UTC)[reply]

A good point of this list is that it is totally non subjective (It would be very difficult to estabilish who are the founders of CG and the VERY notable). There are some absence, but the keeper of the list is very collaborative and willing to integrate/update the list. If the list is considered too long switching to the G' number could be an option. ALoopingIcon 07:34, 25 April 2007 (UTC)[reply]

G number seems like a wider perspective and a higher standard but I'm not sure (i'm not really clear on what a G number is). It's just that as it is, it's hard to judge entries that aren't from that list but may be noteable. Maybe if every researcher was listed with a G number? Adam McCormick 05:29, 26 April 2007 (UTC)[reply]

G number definition (from Ron Fedkiw's site):
A (G)raphics researchers G-number is calculated as the number of papers/books/citations on (G)oogle Scholar that contain more than 200 cites
So if you have written a paper that have been cited by other papers more than 200 times you are eligible to be in the G list. Counting citations is one the basis elements for judging the impact of researcher activity (Bibliometrics, H-index, Impact factor, etc.). The only subjective aspect of this specific G-number is deciding if a given researcher belongs or not to the Computer Graphics community. ALoopingIcon 08:09, 26 April 2007 (UTC)[reply]

I like his reasoning on this, and am threfore of the opinion that the list should come from the G' group. It seems the right size and the reasoning of avoiding hype and short-term publicity is strong. Adam McCormick 15:04, 26 April 2007 (UTC)[reply]

I agree that the list is too long, as is. And I agree that citation count is a nice, objective way to decide who to include. But there is a problem- Computer vision has a lot more researchers and publications, and thus the graphics researchers with a vision slant are favored by these G number calculations. Take Takeo Kanade, the research with the highest G number. He's an amazing researcher, but he's really not much of a computer graphics researcher. He has no SIGGRAPH publications, as far as I know. He's still a noteworthy computer graphics researcher because of all the crossover between vision and graphics and because some of his research arguably is computer graphics (virtual reality, novel viewpoint synthesis), but I don't think many people would name him as the most important computer graphics researcher. Jitendra Malik and Stan Osher are in the same boat. Dmitri Terzopoulos too. Terzopoulos' most cited papers are Vision (IJCV), Graphics (Siggraph), Medical Image Analysis, Vision (PAMI), and Vision (PAMI), etc. I'm not saying these researchers aren't relevant, but the vision people are getting a huge boost in their G number. Marc Levoy is exceptional in that almost all of his publications are actually in computer graphics. Ideally a good objective measure of computer graphics relevance would be the number of paper citations in computer graphics conferences and journals. Switching to G' instead of G actually prunes away a lot of important researchers and leaves a higher percentage of people with citations outside the graphics community, I feel.Jhhays 13:17, 2 June 2007 (UTC)[reply]

It would seem that the best option then would be to take the G group then prune out non-graphics researchers but I'm not sure this can be done neutrally. It would seem the G' would give us notable researchers who have done graphics rather than notable researchers in computer graphics, but it seems like the same is true of the G group except that it happens to include more pure-graphics researchers. Adam McCormick 03:57, 4 June 2007 (UTC)[reply]

I don't think this G number is widely recognized and would probably omit things that belong and include things that don't. Also, some one (who has time) could reorder the names to be alphabetical by last name, not first. Temp432 05:13, 24 September 2007 (UTC)[reply]

After looking at the page a bit more, I'd like to suggest that both the Researchers and the Research Groups lists be moved off onto their own pages. In fact, the list of researchers is somewhat redundant with the Category:Computer_graphics_professionals. This article should focus on what Computer Graphics is because I think that is what most people coming to it are interested in. The links to places and people that do interesting graphics are sort of tangential. Making the list of researchers a separate page helps with the "clutter" issue mentioned above because a long list of names is less objectionable on a page dedicated to listing researchers. Also, the long list of refs generated by all the links to people's home pages is really ugly. Another also, the research groups list should be WP links where possible as several of these groups have WP entries. Temp432 05:46, 24 September 2007 (UTC)[reply]

Ok, as far as G number, it's just a formalization of basic scientific community recognition. The number of other researchers who have quoted or used one's research is the most important validation of a researcher's position. This list is not intended to be any kind of an exhastive list, it should be the MOST influential contributors to the science. I agree that the list is excessive but if properly researched, which I haven't had time or motivation to do, this is the fundamental basis of the notability of the science. I would suggest that this list be moved to another page and perhaps the top ten or so be kept, but detirmining the top ten of such a diverse field is next to impossible. I also agree that they should be orderred by last names. Adam McCormick 07:29, 24 September 2007 (UTC)[reply]

If you move it to another page, then it is probably is okay if the list is long a bit long as it won't be cluttering this article. If you try to do top 10 then there will be fighting at some point about who the top ten are or what measure to use. For example this G list thing counts citations that appear in papers that are in other areas and papers that themselves may be insignifigant. In other words all citations are not created equal. Citeseer uses a different ranking scheme that makes it so a citation in a heavily cited paper is worth more than one in a less cited paper. People also argue about how the age of a citation should factor into the rating, or if it should. Anyhow, I'm sure some people don't like citeseer and have yet another way ranking people. Keep in mind, this G thing was made up by one guy who is an expert on simulation, not reference significance. To illustrate that the G Number is pretty flawed, consider that Stan Osher has a G number of 15 and qualifies for a G',G' ', and G' ' ' number, but I would not have included him in any list of significant graphics researchers. (In actuality I know the guy and I believe that he would not call himself a graphics researcher either.) Similarly, Ingrid Daubechies is the only one with a G' ' ' ' number, and while Ingrid is a world-class expert on wavelets, defiantly not a top computer graphics researcher. This particular problem is that there is no way to say what is or is not a graphics publication. So, for example Gene Golub is being counted due to his fine text on linear algebra. Anyhow, this comment is getting a bit long. I suggest avoiding the problem and doing an inclusive list and maybe taking out a few of the people that are clearly not graphics researchers, but not complaining if someone wants to add a person back in as long as it's reasonable. Temp432 08:08, 24 September 2007 (UTC)[reply]

Yeah, I agree. I mean, I think we all know which of these people are "graphics researchers" and which of them are not... —Preceding unsigned comment added by Trevorgoodchild (talk • contribs) 16:20, 16 October 2007 (UTC)[reply]

A little cleanup

I've hidden the two long lists. I think cleaning up the references is the next logical step. Any Volunteers? Adam McCormick 16:14, 16 October 2007 (UTC)[reply]

Science/Process dichotomy

Ok, so several months ago, we decided that there needs to exist a logical partition between this page and 3D Computer Graphics. I think that it needs to go a step further, and there needs to be a mirrored branching of "study of" articles to match the branching of "process of" articles. The problem I'm seeing is that there are a lot of articles being created that overlap this logical partition which makes the whole heirarchy more dificult to resolve. I'm making fixing this my personal goal and I hope others will help me. Adam McCormick 21:07, 24 October 2007 (UTC)[reply]

62 refs?

Where exactly? --HanzoHattori (talk) 13:44, 13 December 2007 (UTC)[reply]

The references are for computer graphics research groups and notable researchers (which are long, so they were put in hidden lists per another cleanup request). I've put the references in a hidden list as well, since having them expanded is fairly unsightly.Trevorgoodchild (talk) 16:35, 13 December 2007 (UTC)[reply]

Notice

This article focuses on the academic discipline of Computer Graphics rather than applications of computer graphics to movies and video games (as these topics are covered extensively by other Wikipedia pages). It has been modeled after the general article on physics -- the idea being to give a broad overview of the problem, the connection of computer graphics to other disciplines, and some history of the field as a whole. You may want to take a look at the physics article before editing this one to get an idea of what a good overview article looks like.

Specific information about graphics sub-topics is abundant on the Wikipedia and does not need to be repeated here!

Additionally, if you'd like to write something about a specific subtopic (eg. graphics specific to the movie industry or graphics specific to the video game industry) you are enchoraged to contribute to the CGI page, to contribute to one of the video games subpages, or to create a specialized page.

With all due respect, says who? Unsigned injunctions are not taken seriously. If you mean it, sign it. Dhatfield (talk) 16:50, 22 July 2008 (UTC)[reply]

Says me. :) Actually, I did sign this comment originally (over a year ago), but since then the original message has been archived and somehow lost my signature in the process. This notice has more or less guided the spirit of this page since then, though, and many (silent) editors seem to be happy to structure the article in this way. You can look in the second page of the archive if you really want to know more -- Alanbly and myself already had extensive discussions about the purpose of this page. Unfortunately, these same little issues seem to have resurfaced recently without a nod to history... ;) Anyway, good luck with your WikiWars! Trevorgoodchild (talk) 17:15, 22 July 2008 (UTC)[reply]

I notice that in the prior discussion (Archive 2) you stated your viewpoint, another author disagreed and you restated your viewpoint. I think it's time for the 'silent' authors to become vocal again. I'm not a fan of warring and truly hope we can get some consensus. I don't believe that a textual disambiguation of such a prominent and broad term is unreasonable. Dhatfield (talk) 17:38, 22 July 2008 (UTC)[reply]

Policy on Researchers

Per Wikipedia:Notability (people)#Lists_of_people I think we need to consider removing any of the people in the researchers list who could not meet the notability criteria on their own. Preliminarily, this should consist of removing any researcher without articles of their own. I will go and do this unless notability can be asserted for those without articles. Let's give it till the weekend, but then I'm going to go ahead. Adam McCormick (talk) 01:08, 4 March 2008 (UTC)[reply]

Ok, I've waited over a week with no discussion so I'm going ahead and deleting any researchers who do not have wikipedia articles. Adam McCormick (talk) 02:05, 18 March 2008 (UTC)[reply]