Analysis of reported error in Monte Carlo rendered images

Whittle, Joss; Jones, Mark W.; Mantiuk, Rafał

doi:10.1007/s00371-017-1384-7

Cited by 7 publications

(4 citation statements)

References 55 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We analyze the output of the Convolutional Autoencoder in Section 6.1, perform an ablation study in Section 6.2 and evaluate image quality using established metrics like SSIM and Entropy in Section 6.3. SSIM is preferred in our experiments over, e.g., PSNR, since we are analyzing Monte Carlo rendered images [38]. The difference in amount and depth of indirect illumination 54 paths across the test scenes allows us to verify that our network 55 is able to deal with different input data quality seamlessly, and 56 that it is able to extrapolate and adapt to geometry not encoun-57 tered during training.…”

Section: Evaluation and Resultsmentioning

confidence: 97%

Deep radiance caching: Convolutional autoencoders deeper in ray tracing

Jiang

Kainz

2021

Computers & Graphics

View full text Add to dashboard Cite

Rendering realistic images with global illumination is a computationally demanding task and often requires dedicated hardware for feasible runtime. Recent research uses Deep Neural Networks to predict indirect lighting on image level, but such methods are commonly limited to diffuse materials and require training on each scene. We present Deep Radiance Caching (DRC), an efficient variant of Radiance Caching utilizing Convolutional Autoencoders for rendering global illumination. DRC employs a denoising neural network with Radiance Caching to support a wide range of material types, without the requirement of offline pre-computation or training for each scene. This offers high performance CPU rendering for maximum accessibility. Our method has been evaluated on interior scenes, and is able to produce high-quality images within 180 seconds on a single CPU.

show abstract

Section: Evaluation and Resultsmentioning

confidence: 97%

Deep radiance caching: Convolutional autoencoders deeper in ray tracing

Jiang

Kainz

2021

Computers & Graphics

View full text Add to dashboard Cite

show abstract

“…Both qualitative and quantitative metrics were used to benchmark the PT and BDPT algorithms. Regarding the qualitative metric to make the final image quality assessment, Structural Similarity Index Metric (SSIM) (Zhou Wang et al, 2004) is the preferred method for Monte Carlo rendered images (Whittle et al, 2017). We used a tool that computes (dis)similarity (DSSIM) between two or more PNG images using the SSIM algorithm at multiple weighed resolutions (Kornelski, 2020).…”

Section: Testing Methodologymentioning

confidence: 99%

Lift: An Educational Interactive Stochastic Ray Tracing Framework with AI-Accelerated Denoiser

Soares¹,

Pereira²

2021

CSRN

View full text Add to dashboard Cite

Real-time physically based rendering has long been looked at as the holy grail in Computer Graphics. With the introduction of Nvidia RTX-enabled GPUs family, light transport simulations under real-time constraint started to look like a reality. This paper presents Lift, an educational framework written in C++ that explores the RTX hardware pipeline by using the low-level Vulkan API and its Ray Tracing extension, recently made available by Khronos Group. Furthermore, to accomplish low variance rendered images, we integrated the AI-based denoiser available from the Nvidia ́s OptiX framework. Lift’s development arose primarily in the context of the graduate 3D Programming course taught at Instituto Superior Técnico and Master Theses focused on Real-Time Ray Trac- ing and provides the foundations for laboratory assignments and projects development. The platform aims to make easier students to learn and to develop, by programming the shaders of the RT pipeline, their physically-based ren- dering approaches and to compare them with the built-in progressive unidirectional and bidirectional path tracers. The GUI allows a user to specify camera settings and navigation speed, to select the input scene as well as the rendering method, to define the number of samples per pixel and the path length as well as to denoise the generated image either every frame or just the final frame. Statistics related with the timings, image resolution and total number of accumulated samples are provided too. Such platform will teach that nowadays physically-accurate images can be rendered in real-time under different lighting conditions and how well a denoiser can reconstruct images rendered with just one sample per pixel.

show abstract

“…Subr and Kautz [SK13] analyzed the error caused by various Monte Carlo (MC) sampling patterns via Fourier analysis, but their conclusions do not generalize. Whittle et al [WJM17] provided an extensive overview of error metrics for images, and they examined the influence of poor references for computing those measures. However, they did not investigate their variance, or sensitivity to outliers.…”

Section: Introductionmentioning

confidence: 99%

“…Our reference solutions were computed using a cluster and millions of samples per pixel ‐ several orders of magnitude more than were used for 〈 I 〉 N . To put this into perspective, Whittle et al recommended to use reference images with at least an order of magnitude more samples [WJM17]. If the T , 4 T , and 16 T curves overlap, then the algorithm behaves like in that frequency and rendering budget range.…”

Section: Introductionmentioning

confidence: 99%

Quantifying the Error of Light Transport Algorithms

Celarek

Jakob

Wimmer

et al. 2019

Computer Graphics Forum

View full text Add to dashboard Cite

This paper proposes a new methodology for measuring the error of unbiased physically based rendering algorithms. The current state of the art includes mean squared error (MSE) based metrics and visual comparisons of equal‐time renderings of competing algorithms. Neither is satisfying as MSE does not describe behavior and can exhibit significant variance, and visual comparisons are inherently subjective. Our contribution is two‐fold: First, we propose to compute many short renderings instead of a single long run and use the short renderings to estimate MSE expectation and variance as well as per‐pixel standard deviation. An algorithm that achieves good results in most runs, but with occasional outliers is essentially unreliable, which we wish to quantify numerically. We use per‐pixel standard deviation to identify problematic lighting effects of rendering algorithms. The second contribution is the error spectrum ensemble (ESE), a tool for measuring the distribution of error over frequencies. The ESE serves two purposes: It reveals correlation between pixels and can be used to detect outliers, which offset the amount of error substantially.

show abstract

Analysis of reported error in Monte Carlo rendered images

Cited by 7 publications

References 55 publications

Deep radiance caching: Convolutional autoencoders deeper in ray tracing

Deep radiance caching: Convolutional autoencoders deeper in ray tracing

Lift: An Educational Interactive Stochastic Ray Tracing Framework with AI-Accelerated Denoiser

Quantifying the Error of Light Transport Algorithms

Contact Info

Product

Resources

About