SBMCD (Sample-based Monte Carlo Denoising using a Kernel-Splatting Network)

Discussion related to the Engine functionality, implementations and API.
Odilkhan Yakubov
Posts: 34
Joined: Fri Jan 26, 2018 10:07 pm
Location: Tashkent, Uzbekistan

SBMCD (Sample-based Monte Carlo Denoising using a Kernel-Splatting Network)

Post by Odilkhan Yakubov » Thu Jun 13, 2019 6:14 pm

Hi there. When I surfing on the internet I found a brend-new denoising algorithm which is called "Sample-based Monte Carlo Denoising using a Kernel-Splatting Network" aka SBMCD.

Original info from the authors:
Image
Denoising has proven to be useful to efficiently generate high-quality Monte Carlo renderings. Traditional pixel-based denoisers exploit summary statistics of a pixel's sample distributions, which discards much of the samples' information and limits their denoising power. On the other hand, sample-based techniques tend to be slow and have difficulties handling general transport scenarios. We present the first convolutional network that can learn to denoise Monte Carlo renderings directly from the samples. Learning the mapping between samples and images creates new challenges for the network architecture design: the order of the samples is arbitrary, and they should be treated in a permutation invariant manner. To address these challenges, we develop a novel kernel-predicting architecture that splats individual samples onto nearby pixels. Splatting is a natural solution to situations such as motion blur, depth-of-field and many light transport paths, where it is easier to predict which pixels a sample contributes to, rather than a gather approach that needs to figure out, for each pixel, which samples (or nearby pixels) are relevant. Compared to previous state-of-the-art methods, ours is robust to the severe noise of low-sample count images (e.g. 8 samples per pixel) and yields higher-quality results both visually and numerically. Our approach retains the generality and efficiency of pixel-space methods while enjoying the expressiveness and accuracy of the more complex sample-based approaches.

DOWNLOAD:https://groups.csail.mit.edu/graphics/r ... oising.pdf
Note: I already asked Dade and Simon on working on this algorithm to implementation into LuxCore.
Allah bless us!

User avatar
Dade
Developer
Developer
Posts: 2488
Joined: Mon Dec 04, 2017 8:36 pm

Re: SBMCD (Sample-based Monte Carlo Denoising using a Kernel-Splatting Network)

Post by Dade » Fri Jun 14, 2019 8:43 am

Thanks, it looks interesting however Intel Oidn works well and it is one less problem to solve for us so, given our lack of resources, it is hard to consider anything outside an already available solution that works well and it is easy to integrate like Oidn. It is the same with Intel Embree.

If someone else want to work on this topic, sure, otherwise I will hardly find the time to look into it.
Support LuxCoreRender project with salts and bounties

Odilkhan Yakubov
Posts: 34
Joined: Fri Jan 26, 2018 10:07 pm
Location: Tashkent, Uzbekistan

Re: SBMCD (Sample-based Monte Carlo Denoising using a Kernel-Splatting Network)

Post by Odilkhan Yakubov » Fri Jun 14, 2019 9:33 am

Thank you Dade. I'm considering into it. It will be cool if someone implement this feature to the engine.
Allah bless us!

User avatar
lacilaci
Donor
Donor
Posts: 1093
Joined: Fri May 04, 2018 5:16 am

Re: SBMCD (Sample-based Monte Carlo Denoising using a Kernel-Splatting Network)

Post by lacilaci » Fri Jun 14, 2019 9:48 am

Is there any benefit to this agains intel's OIDN? Results aren't that impressive

Odilkhan Yakubov
Posts: 34
Joined: Fri Jan 26, 2018 10:07 pm
Location: Tashkent, Uzbekistan

Re: SBMCD (Sample-based Monte Carlo Denoising using a Kernel-Splatting Network)

Post by Odilkhan Yakubov » Fri Jun 14, 2019 10:19 am

Why? Is it so unuseful thing, are you think so?
Allah bless us!

Odilkhan Yakubov
Posts: 34
Joined: Fri Jan 26, 2018 10:07 pm
Location: Tashkent, Uzbekistan

Re: SBMCD (Sample-based Monte Carlo Denoising using a Kernel-Splatting Network)

Post by Odilkhan Yakubov » Fri Jun 14, 2019 10:42 am

@Dade, can you try to explain to us this algo in the example?
Allah bless us!

User avatar
Sharlybg
Donor
Donor
Posts: 1325
Joined: Mon Dec 04, 2017 10:11 pm
Location: Ivory Coast

Re: SBMCD (Sample-based Monte Carlo Denoising using a Kernel-Splatting Network)

Post by Sharlybg » Fri Jun 14, 2019 10:43 am

Why? Is it so unuseful thing, are you think so?
it isn't that people think it as a bad feature or a super good feature. It is simply that we all need a reference point.

How it compare against current working solution.past month the lux team change from BCD denoiser To intel OIDN because of better performance (memory / speed / quality).And before that we also tried to introduce Nvidia AI denoiser.

This community is far from being closed as Dade said human ressources here is the main issue.Dev have lot of work on the hands.
We are asking just to have better idea of the feature and improvement SBMCD bring on the table. just that.
please don't be offended ;)
Support LuxCoreRender project with salts and bounties

Portfolio : https://www.behance.net/DRAVIA

User avatar
FarbigeWelt
Donor
Donor
Posts: 654
Joined: Sun Jul 01, 2018 12:07 pm
Location: Switzerland
Contact:

Re: SBMCD (Sample-based Monte Carlo Denoising using a Kernel-Splatting Network)

Post by FarbigeWelt » Fri Jun 14, 2019 10:48 am

lacilaci wrote:
Fri Jun 14, 2019 9:48 am
Results aren't that impressive
Are you sure? Did you look at the pictures on p. 125:11 of the paper linked on main post?
160.8 | 42.8 (10.7) Gfp / Windows 10 Pro, intel i7 4770K@3.5, 32 GB | AMD R9 290x+R9 390x, 4 GB
17.3 | 19.0 ( 4.7) Gfp / macOS X 13.6, iMac 27'', 2010, intel i7 870@2.93, 24 GB | ATI Radeon HD 5750, 1 GB
#luxcorerender | Gfp = SFFT Gflops

User avatar
FarbigeWelt
Donor
Donor
Posts: 654
Joined: Sun Jul 01, 2018 12:07 pm
Location: Switzerland
Contact:

Re: SBMCD (Sample-based Monte Carlo Denoising using a Kernel-Splatting Network)

Post by FarbigeWelt » Fri Jun 14, 2019 10:56 am

Sharlybg wrote:
Fri Jun 14, 2019 10:43 am
We are asking just to have better idea of the feature and improvement SBMCD bring on the table.
What does SBMCD mean? What is its purpose?
160.8 | 42.8 (10.7) Gfp / Windows 10 Pro, intel i7 4770K@3.5, 32 GB | AMD R9 290x+R9 390x, 4 GB
17.3 | 19.0 ( 4.7) Gfp / macOS X 13.6, iMac 27'', 2010, intel i7 870@2.93, 24 GB | ATI Radeon HD 5750, 1 GB
#luxcorerender | Gfp = SFFT Gflops

User avatar
B.Y.O.B.
Developer
Developer
Posts: 2648
Joined: Mon Dec 04, 2017 10:08 pm
Location: Germany
Contact:

Re: SBMCD (Sample-based Monte Carlo Denoising using a Kernel-Splatting Network)

Post by B.Y.O.B. » Fri Jun 14, 2019 11:02 am

I only skimmed through the paper quickly but one interesting feature I noticed is that this denoiser doesn't require a clean albedo AOV.
Support LuxCoreRender project with salts and bounties

Post Reply