Vassillen Chizhov
Iliyan Georgiev
Skip Abstract Section
Skip Supplemental Material SectionAbstract
Synthesizing realistic images involves computing high-dimensional light-transport integrals. In practice, these integrals are numerically estimated via Monte Carlo integration. The error of this estimation manifests itself as conspicuous aliasing or noise. To ameliorate such artifacts and improve image fidelity, we propose a perception-oriented framework to optimize the error of Monte Carlo rendering. We leverage models based on human perception from the halftoning literature. The result is an optimization problem whose solution distributes the error as visually pleasing blue noise in image space. To find solutions, we present a set of algorithms that provide varying trade-offs between quality and speed, showing substantial improvements over prior state of the art. We perform evaluations using quantitative and error metrics and provide extensive supplemental material to demonstrate the perceptual improvements achieved by our methods.
Supplemental Material
Available for Download
zip
chizhov.zip (167.4 MB)
Supplemental movie, appendix, image and software files for, Perceptual Error Optimization for Monte Carlo Rendering
- [1] . 2020. Screen-space blue-noise diffusion of Monte Carlo sampling error via hierarchical ordering of pixels. ACM Trans. Graph. (Proc. SIGGRAPH Asia) 39, 6, Article 244 (2020).
DOI: Google Scholar
Digital Library
- [2] . 1976. Random quasi-periodic halftone process. J. Opt. Soc. Am. 66, 9 (
Sep. 1976), 909–917.DOI: Google ScholarCross Ref
- [3] . 1992. Model-based halftoning using direct binary search. In Human Vision, Visual Processing, and Digital Display III, (Ed.), Vol. 1666. International Society for Optics and Photonics, SPIE, 96–108.
DOI: Google ScholarCross Ref - [4] . 1989. Error diffusion coding for A/D conversion. IEEE Trans. Circ. Syst. 36, 9 (1989), 1175–1186.Google ScholarCross Ref
- [5] . 2017. The academy color encoding system (ACES): A professional color-management framework for production, post-production and archival of still and motion pictures. J. Imag. 3 (Sep. 2017), 40.
DOI: Google ScholarCross Ref - [6] . 1999. Contrast Sensitivity of the Human Eye and Its Effects on Image Quality. The International Society for Optical Engineering (SPIE).Google ScholarCross Ref
- [7] . 1973. An optimum method for two-level rendition of continuous-tone pictures. In Proceedings of IEEE International Conference on Communications, Conference Record, Vol. 26. 11–15.Google Scholar
- [8] . 2021. Lessons learned and improvements when building screen-space samplers with blue-noise error distribution. In Proceedings of the ACM SIGGRAPH Talks.
DOI: Google ScholarDigital Library - [9] . 1995. A frequency based ray tracer. In Proceedings of the ACM SIGGRAPH. 409–418.Google ScholarDigital Library
- [10] . 1998. A perceptually based adaptive sampling algorithm. In Proceedings of the ACM SIGGRAPH. 299–309.Google ScholarDigital Library
- [11] . 2019. Quantifying the error of light transport algorithms. Comput. Graph. Forum 38, 4 (2019), 111–121.
DOI: Google ScholarCross Ref - [12] . 2017. Interactive reconstruction of Monte Carlo image sequences using a recurrent denoising autoencoder. ACM Trans. Graph. 36, 4, Article 98 (2017).Google ScholarDigital Library
- [13] . 2009. Structure-aware error diffusion. ACM Trans. Graph. (Proc. SIGGRAPH Asia) 28, 5, Article 162 (2009).
DOI: Google ScholarDigital Library - [14] . 1987. Subroutine for the generation of a two dimensional human visual contrast sensitivity function. Technical Report 233203Y.Google Scholar
- [15] . 1993. The visible differences predictor: An algorithm for the assessment of image fidelity. In Digital Image and Human Vision, (Ed.). MIT Press, Cambridge, MA, 179–206.Google ScholarDigital Library
- [16] . 1991. A simple parametric model of the human ocular modulation transfer function. Ophthalmol. Physiol. Optics 11 (1991), 91–93.Google ScholarCross Ref
- [17] . 1996. A model of visual adaptation for realistic image synthesis. In Proceedings of the ACM SIGGRAPH. 249–258.
DOI: Google ScholarDigital Library - [18] . 1976. An adaptive algorithm for spatial greyscale. Proceedings of the Society for Information Display 17, 2 (1976), 75–77.Google Scholar
- [19] . 2016. Blue-noise dithered sampling. In Proceedings of the ACM SIGGRAPH Talks. Article 35.
DOI: Google ScholarDigital Library - [20] . 2006. Alpha stable human visual system models for digital halftoning. In Electronic Imaging.Google Scholar
- [21] . 2019. Distributing Monte Carlo errors as a blue noise in screen space by permuting pixel seeds between frames. Comput. Graph. Forum (2019).
DOI: Google ScholarCross Ref - [22] . 2019. A low-discrepancy sampler that distributes Monte Carlo errors as a blue noise in screen space. In Proceedings of the SIGGRAPHTalks. Retrieved from https://hal.archives-ouvertes.fr/hal-02150657.Google ScholarDigital Library
- [23] . 1994. Abstract Harmonic Analysis: Volume I Structure of Topological Groups Integration Theory Group Representations. Springer, New York, NY.Google Scholar
- [24] . 2008. Reinstating Floyd-Steinberg: Improved metrics for quality assessment of error diffusion algorithms, Vol. 5099. 38–45.
DOI: Google ScholarDigital Library - [25] . 2018. Intel Open Image Denoise. Retrieved from https://www.openimagedenoise.org.Google Scholar
- [26] . 1986. The rendering equation. In Proceedings of the ACM SIGGRAPH, Vol. 20. 143–150.Google ScholarDigital Library
- [27] . 2019. DeepFovea: Neural reconstruction for foveated rendering and video compression using learned statistics of natural videos. ACM Trans. Graph. 38, 6, Article 212 (2019).Google ScholarDigital Library
- [28] . 2014. A GPU implementation of clipping-free halftoning using the direct binary search. In Algorithms and Architectures for Parallel Processing, , , , , , , , , , and (Eds.). Springer International Publishing, Cham, 57–70.Google ScholarCross Ref
- [29] . 1974. Uniform Distribution of Sequences. Wiley, New York, NY.Google Scholar
- [30] . 2018. Deep adaptive sampling for low sample count rendering. Comput. Graph. Forum 37, 4 (2018), 35–44.Google ScholarCross Ref
- [31] . 2007. Modern Digital Halftoning, 2nd ed. CRC Press, Boca Raton, FL.Google Scholar
- [32] . 1990. Scale-space for discrete signals. IEEE Trans. Pattern Anal. Mach. Intell. 12, 3 (1990), 234–254.Google ScholarDigital Library
- [33] . 1995. A visual discrimination model for imaging system design and evaluation. In Vision Models for Target Detection and Recognition, (Ed.). World Scientific, Singapore, 245–283.Google ScholarCross Ref
- [34] . 1974. The effects of a visual fidelity criterion of the encoding of images. IEEE Trans. Info. Theory 20, 4 (1974), 525–536.
DOI: Google ScholarDigital Library - [35] . 2005. Predicting visible differences in high dynamic range images: Model and its calibration. In Human Vision and Electronic Imaging X, Vol. 5666. SPIE, 204–214.Google Scholar
- [36] . 2011. HDR-VDP-2: A calibrated visual metric for visibility and quality predictions in all luminance conditions. ACM Trans. Graph. (Proc. SIGGRAPH) 30, 4 (2011).Google ScholarDigital Library
- [37] . 2003. Parallel digital halftoning by error-diffusion. In Proceedings of the Paris C. Kanellakis Memorial Workshop on Principles of Computing and Knowledge. 35–41.
DOI: Google ScholarDigital Library - [38] . 1991. Spectrally optimal sampling for distribution ray tracing. In Proceedings of the ACM SIGGRAPH, Vol. 25. 157–164.
DOI: Google ScholarDigital Library - [39] . 1991. Digital halftoning using a blue noise mask. In Proceedings of International Conference on Acoustics, Speech, and Signal Processing. 2809–2812, vol. 4.
DOI: Google ScholarCross Ref - [40] . 1985. The contrast sensitivity of human colour vision to red-green and blue-yellow chromatic gratings. J. Physiol. 359 (1985), 381–400.Google ScholarCross Ref
- [41] . 1998. The visible differences predictor: Applications to global illumination problems. In Proceedings of the Conference on Rendering Techniques, and (Eds.). Springer, Vienna, 223– 236.Google ScholarCross Ref
- [42] . 1984. Visibility of halftone dot textures. IEEE Trans. Syst., Man, Cybernet. SMC-14, 6 (1984), 920–924.Google ScholarCross Ref
- [43] . 1992. Random Number Generation and Quasi-Monte Carlo Methods. Society for Industrial and Applied Mathematics, Philadelphia, PA.Google ScholarCross Ref
- [44] . 2001. A simple and efficient error-diffusion algorithm. In Proceedings of the ACM SIGGRAPH.
DOI: Google ScholarDigital Library - [45] . 2008. Structure-aware halftoning. ACM Trans. Graph. (Proc. ACM SIGGRAPH) 27, 4, Article 89 (2008).
DOI: Google ScholarDigital Library - [46] . 1999. Least-squares model-based halftoning. IEEE Trans. Image Process. 8, 8 (
Aug. 1999), 1102–1116.DOI: Google ScholarDigital Library - [47] . 1991. Image invariance with changes in size: The role of peripheral contrast thresholds. J. Opt. Soc. Am. A 8, 11 (1991), 1762–1774.Google ScholarCross Ref
- [48] . 2016. Physically Based Rendering: From Theory To Implementation (3rd ed.). Morgan Kaufmann.Google Scholar
- [49] . 1999. A perceptually based physical error metric for realistic image synthesis. In Proceedings of the ACM SIGGRAPH. 73–82.Google ScholarDigital Library
- [50] . 2005. 5.5-lossy image compression: JPEG and JPEG2000 standards. In Handbook of Image and Video Processing), 2nd ed., (Ed.). Academic Press, San Diego, CA, 709–XXII.Google Scholar
- [51] . 1981. Probability summation and regional variation in contrast sensitivity across the visual field. Vision Res. 21, 3 (1981), 409–418.Google ScholarCross Ref
- [52] . 2019. Analysis of sample correlations for Monte Carlo rendering. Comput. Graph Form. (Proc. EGSR) 38, 2 (2019).Google Scholar
- [53] . 2011. Comparative neurophysiology of spatial luminance contrast sensitivity. Psychol. Neurosci. 4 (
06 2011), 29–48.Google ScholarCross Ref - [54] . 1997. Methods for generating blue-noise dither matrices for digital halftoning. J. Electr. Imag. 6, 2 (1997), 208–230.
DOI: Google ScholarCross Ref - [55] . 1991. Design of minimum visual modulation halftone patterns. IEEE Trans. Syst., Man, Cybernet. 21, 1 (
Jan. 1991), 33–38.DOI: Google ScholarCross Ref - [56] . 1987. Digital Halftoning. MIT Press, Cambridge, MA.
87003182 Google ScholarDigital Library - [57] . 1993. Void-and-cluster method for dither array generation. In Human Vision, Visual Processing, and Digital Display IV, vol. 1913, Jan P. Allebach and Bernice E. Rogowitz (Eds.). International Society for Optics and Photonics, 332–343.
DOI: Google ScholarCross Ref - [58] . 2000. Using the visual differences predictor to improve performance of progressive global illumination computation. ACM Trans. Graph. 19, 2 (2000), 122– 161.Google ScholarDigital Library
- [59] . 1986. The eye as an optical instrument. In Handbook of Perception and Human Performance: 1. Sensory Processes and Perception, , , and (Eds.). Wiley, New York, 4.1–4.20.Google Scholar
- [60] . 2020. Spatio-chromatic contrast sensitivity under mesopic and photopic light levels. J. Vision 20, 4 (Apr. 2020), 23–23.Google ScholarCross Ref
- [61] . 2015. Recent advances in adaptive sampling and reconstruction for Monte Carlo rendering. Comput. Graph. Forum 34, 2 (2015), 667– 681.Google ScholarCross Ref
Index Terms
- Perceptual Error Optimization for Monte Carlo Rendering
Recommendations
Perceptual error optimization for Monte Carlo animation rendering
SA '23: SIGGRAPH Asia 2023 Conference PapersIndependently estimating pixel values in Monte Carlo rendering results in a perceptually sub-optimal white-noise distribution of error in image space. Recent works have shown that perceptual fidelity can be improved significantly by distributing pixel ...
Screen-space blue-noise diffusion of monte carlo sampling error via hierarchical ordering of pixels
We present a novel technique for diffusing Monte Carlo sampling error as a blue noise in screen space. We show that automatic diffusion of sampling error can be achieved by ordering the pixels in a way that preserves locality, such as Morton's Z-...
Improving Monte Carlo Efficiency by Increasing Variance
This paper compares the performances of two well-known Monte Carlo procedures for estimating an unknown quantity as the size of the problem grows. One method based on the standard Monte Carlo approach generates K i.i.d. data points. The other derives ...
Comments