PDA

View Full Version : SIGGRAPH 2005 papers


donVITO
05-13-2005, 06:26 PM
This page is maintained by Ke-Sen Huang and Tim Rowley,
show links to some SIGGRAPH 2005 papers:

http://www.cs.brown.edu/~tor/sig2005.html


The Fluid Simulation section is superb.

SoLiTuDe
05-30-2005, 02:43 AM
thanks for the link! I'm sure i'll be up for hours reading, and feeling more and more insecure :)

ngrava
05-30-2005, 08:29 AM
OMG!! Check out the mesh based IK. It's completely amazing! I never would have thought it would be possible to do this if I had not seen it. I've thought about deformation caches before but never like this. This opens up a whole new area for character animation that's not limited by skeletal systems at all. Basically, no rigging. Well, the rigging stage is replaced by a heavy modeling stage but you get exactly what you want.

I'm totally blown away...

-=GB=-

mummey
05-30-2005, 03:28 PM
Unfortunately every paper I'm interested in is not up yet :-P

LangFox
05-31-2005, 09:04 AM
Good papers,thank you! :bounce:

funfun
06-11-2005, 11:06 AM
"Light Diffusion in Multi-Layered Translucent Materials" (http://graphics.ucsd.edu/papers/layered) by Henrik Wann Jensen

http://graphics.ucsd.edu/papers/layered/face.jpg
http://graphics.ucsd.edu/papers/layered/face_nose.jpg

playmesumch00ns
06-13-2005, 09:44 AM
That paper's all well and good but if you look at the before/after images you notice there's not a hell of a lot of difference!

BigSky
06-18-2005, 10:19 AM
Lots of cool stuff, as ever. My favorite...As rigid as possible mesh deformations (http://www-ui.is.s.u-tokyo.ac.jp/~takeo/research/rigid/index.html) why? because my kids can use it, so animation becomes ubiquitous. :thumbsup:

donVITO
06-21-2005, 05:06 PM
Perhaps anybody has not seen it:

http://www.cs.brown.edu/~tor (older papers)

N4x0
07-02-2005, 11:13 AM
Don't you know real flow 3.0???

Real Flow (www.realflow.com)

DaKrunch
07-11-2005, 03:36 AM
That paper's all well and good but if you look at the before/after images you notice there's not a hell of a lot of difference!


It's actually very good though a bit intricated, even if i am not implementing it on human skin.

The result with painted jade are surprisingly good for a single surface.

playmesumch00ns
07-11-2005, 08:58 AM
Are you actually implementing it? What are you going to use it for?

DaKrunch
07-11-2005, 11:27 PM
PRman (and mayaman for rendering with maya).

still lots to do though

playmesumch00ns
07-12-2005, 08:09 AM
For skin? Or for something else? Penguins?

I ask as I've moved away from the full dipole diffusion approximation for skin. I've found a multilayered, windowed function looking only at local points to be better for simulating that since trying to account for blocking geometry quickly becomes slow.

DaKrunch
07-12-2005, 11:49 PM
Can't say much...
it's not for the penguins though, it's for elephant seals.
i used the 2-pass full dipole diffusion mixed with single scatter -similar to jensen 2002- and it works cool for ice, but it doesn't look great for skin.

cheers

playmesumch00ns
07-13-2005, 08:32 AM
Yeah I've found it to be great for highly scattering materials. Can't say much myself, but we're using my implementation for doing a translucent goo-thing. The irradiance-caching octree concept is probably the best thing about it.

For skin I prefer using a custom function with a windowed lookup. I can evaluate this relatively quickly which allows me to do a multi-layered thing similar to the misss_fast_skin shader from mental images.

Banty
07-20-2005, 10:43 AM
I post the link of my Siggraph's poster which talks about Sub Surface Scattering (it's not physically based):
http://www.geocities.com/frabante/

funfun
07-20-2005, 02:22 PM
Yeah I've found it to be great for highly scattering materials. Can't say much myself, but we're using my implementation for doing a translucent goo-thing. The irradiance-caching octree concept is probably the best thing about it.

For skin I prefer using a custom function with a windowed lookup. I can evaluate this relatively quickly which allows me to do a multi-layered thing similar to the misss_fast_skin shader from mental images.

What is windowed lookup? can you post any image with your SSS implementaion?

DaKrunch
07-20-2005, 11:39 PM
Fra you are everywhere :|

good luck with your poster

playmesumch00ns
07-21-2005, 11:37 AM
What is windowed lookup? can you post any image with your SSS implementaion?

It just means I ignore all points outside a certain distance from the point to be shaded. As long as your falloff function reaches zero at the edge of the window, it doesn't create any artifacts.

I can't post any images tho, they don't belong to me :)

rendermaniac
07-21-2005, 04:26 PM
Is that similar to Christophe Hery's local bounce illumination Stupid RenderMan Trick? How exactly does misss_fast_skin work anyway? I'm guessing it is storing the illumination in the lightmap and then blurring it over a certain radius in the render??

Simon

playmesumch00ns
07-22-2005, 10:52 AM
Is that similar to Christophe Hery's local bounce illumination Stupid RenderMan Trick?

Not really, it's more like what Hery describes using a pointcloud in the "Implementing a Skin BSSDRF..." course from a couple of years ago, but just looking at points within a certain radius.

misss_fast_skin works with lightmaps, yes. Not exacty what kind of kernel it uses tho.

CGTalk Moderation
07-22-2005, 10:52 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.