Results 1 to 16 of 16
  1. #1

    Deep Scattering: Rendering Atmospheric Clouds with Radiance-Predicting Neural Network

    New paper from Disney on Deep Scattering supported by neural networks:

    http://www.drz.disneyresearch.com/~j...eepScattering/

    Stunning results, reduced render time from hours to seconds/minutes.

    Two Minute Papers on this topic:

    Attached Images Attached Images



  2. #2
    Originally Posted by cpurender View Post
    New paper from Disney on Deep Scattering supported by neural networks:

    http://www.drz.disneyresearch.com/~j...eepScattering/

    Stunning results, reduced render time from hours to seconds/minutes.

    Two Minute Papers on this topic:


    verry interesting, makes one wonder how much other parts might benefit from neural networks.
    As light calculations seam to be the most difficult part of rendering



  3. #3
    Pretty cool



  4. #4
    Member m9105826's Avatar
    Join Date
    Dec 2007
    Location
    Fairfax, VA
    Posts
    4,205
    As cool as predictive algorithms for rendering and image manipulation are, I can pretty much guarantee that we're not going to see any FNN, CNN, or PNN code in Blender any time in the near future.
    Long time 3D artist and member of the official Cycles Artists Module
    https://www.youtube.com/user/m9105826 - Training, other stuff. Like and subscribe for more!
    Follow me on Twitter: @mattheimlich or on my blog



  5. #5
    Originally Posted by m9105826 View Post
    As cool as predictive algorithms for rendering and image manipulation are, I can pretty much guarantee that we're not going to see any FNN, CNN, or PNN code in Blender any time in the near future.
    Just out of curiosity, why do you think so?



  6. #6
    Member m9105826's Avatar
    Join Date
    Dec 2007
    Location
    Fairfax, VA
    Posts
    4,205
    Originally Posted by Dantus View Post
    Just out of curiosity, why do you think so?
    Because NNs require either:

    A.) Lots of training

    or

    B.) Lots of cached training data

    Not to mention coders who are familiar with writing NNs, which aren't exactly a dime a dozen.
    Long time 3D artist and member of the official Cycles Artists Module
    https://www.youtube.com/user/m9105826 - Training, other stuff. Like and subscribe for more!
    Follow me on Twitter: @mattheimlich or on my blog



  7. #7
    Member Ace Dragon's Avatar
    Join Date
    Feb 2006
    Location
    Wichita Kansas (USA)
    Posts
    28,343
    Originally Posted by m9105826 View Post
    Because NNs require either:

    A.) Lots of training

    or

    B.) Lots of cached training data

    Not to mention coders who are familiar with writing NNs, which aren't exactly a dime a dozen.
    Not to mention that the training data shown in the video appears to mainly cover cumulus clouds, so I'm left to wonder if you will need many times more images to create accurate depictions of the dozens of other cloud types that are out there (especially if you add the clouds seen on other planets).

    Also to note, the specific material the learning algorithm is designed for (it would get a bit trickier to expand it to general volumetric rendering).
    Sweet Dragon dreams, lovely Dragon kisses, gorgeous Dragon hugs. How sweet would life be to romp with Dragons, teasing you with their fire and you being in their games, perhaps they can even turn you into one as well.
    Adventures in Cycles; My official sketchbook



  8. #8
    Originally Posted by Ace Dragon View Post
    Not to mention that the training data shown in the video appears to mainly cover cumulus clouds, so I'm left to wonder if you will need many times more images to create accurate depictions of the dozens of other cloud types that are out there (especially if you add the clouds seen on other planets).

    Also to note, the specific material the learning algorithm is designed for (it would get a bit trickier to expand it to general volumetric rendering).
    My question was meant in a more general way. For this particular topic, I would be surprised if a generally applicable solution would be found in the near future, something that works for all sorts of clouds and other volumetrics like fog. I would not believe it until someone rendered Cosmos Laundromat successfully with it .



  9. #9
    If it is trained for cumulus clouds, it could just as well be trained for other clouds.
    While training takes some time, using a NN doesnt take much time. And its not that training takes month's for NN, usually the cases i've trained NN its more something like hours or days, which is not a unusual training time, if i compare it to animation rendering.
    Oh and and those GPU's are also excellent for NN



  10. #10
    Member
    Join Date
    Sep 2012
    Posts
    2,771
    Now, is fog a cloud? Basically it is. What happens when camera enters the cloud? Visualization thus depends on POV (proximity).
    Haven't read the paper, but does it consider it.
    Last edited by burnin; 16-Nov-17 at 05:06.



  11. #11
    Originally Posted by Geographic View Post
    If it is trained for cumulus clouds, it could just as well be trained for other clouds.
    While training takes some time, using a NN doesnt take much time. And its not that training takes month's for NN, usually the cases i've trained NN its more something like hours or days, which is not a unusual training time, if i compare it to animation rendering.
    Oh and and those GPU's are also excellent for NN
    In this particular case, creating the training data is the bottleneck. Getting a coverage for volumetrics in general would likely take a lot more.



  12. #12
    Originally Posted by burnin View Post
    What happens when camera enters the cloud? Visualization thus depends on POV (proximity).
    Haven't read the paper, but does it consider it.
    Please read the paper, the NN just predicts the radiance for various shadings, it has nothing to do with the image or camera setup.



  13. #13
    Member
    Join Date
    Sep 2012
    Posts
    2,771
    Aha... "Monkey see, monkey do."

    Thank you.



  14. #14
    Member Marc Driftmeyer's Avatar
    Join Date
    Sep 2014
    Location
    Pacific Northwest WA
    Posts
    44
    From my limited exposure to Blender's particle system it is an absolute pig on resources and render times.



  15. #15
    Member Ace Dragon's Avatar
    Join Date
    Feb 2006
    Location
    Wichita Kansas (USA)
    Posts
    28,343
    Originally Posted by Marc Driftmeyer View Post
    From my limited exposure to Blender's particle system it is an absolute pig on resources and render times.
    Rendering of Blender particle systems in Cycles is often done with the point density texture node.

    The reason why renders take so long is because Cycles has no way of varying the sample counts based on which areas are of interest (which would be possible to do if Cycles gets native OpenVDB rendering support). The solution to the render time issue is known, we just need someone to implement it.
    Sweet Dragon dreams, lovely Dragon kisses, gorgeous Dragon hugs. How sweet would life be to romp with Dragons, teasing you with their fire and you being in their games, perhaps they can even turn you into one as well.
    Adventures in Cycles; My official sketchbook



  16. #16
    Member Marc Driftmeyer's Avatar
    Join Date
    Sep 2014
    Location
    Pacific Northwest WA
    Posts
    44
    Originally Posted by Ace Dragon View Post
    Not to mention that the training data shown in the video appears to mainly cover cumulus clouds, so I'm left to wonder if you will need many times more images to create accurate depictions of the dozens of other cloud types that are out there (especially if you add the clouds seen on other planets).

    Also to note, the specific material the learning algorithm is designed for (it would get a bit trickier to expand it to general volumetric rendering).
    Holy crap, it's Disney and PIXAR. In three years this conversation of complaining will be in the past.



Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •