Primary Sample Space Path Guiding

Jerry Jinfeng Guo; Pablo Bauszat; Jacco Bikker; Elmar Eisemann

doi:10.2312/sre.20181174

Primary Sample Space Path Guiding

Jerry Jinfeng Guo, Pablo Bauszat, Jacco Bikker, Elmar Eisemann

Comp Graphics & Visualisation

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

Abstract

We present a scheme for unbiased path guiding. Different from existing methods that focus on constructing structures in spatial-directional domain, we work in primary sample space. We collect records containing a few dimensions of random numbers as well as the luminance that the resulting path contributes. A multiple dimensional structure is built with collected information. After this, random numbers are drawn from this structure and is used to feed the path tracer. Using this scheme, we are able to work completely outside the rendering kernel. We demonstrate that our method is practical and can be efficient. We manage to reduce variance and reduce zero radiance paths by only working in the primary sample space.

Original language	English
Title of host publication	Eurographics Symposium on Rendering - Experimental Ideas and Implementations
Editors	Wenzel Jakob, Toshiya Hachisuka
Publisher	Eurographics Association
Pages	73-82
Number of pages	10
ISBN (Electronic)	978-3-03868-068-0
DOIs	https://doi.org/10.2312/sre.20181174
Publication status	Published - 2018
Event	EGSR 2018: 29th Eurographics Symposium on Rendering - Karlsruhe Institute of Technology, Karlsruhe, Germany Duration: 2 Jul 2018 → 4 Jul 2018 Conference number: 29 http://cg.ivd.kit.edu/egsr18/

Conference

Conference	EGSR 2018
Country	Germany
City	Karlsruhe
Period	2/07/18 → 4/07/18
Internet address	http://cg.ivd.kit.edu/egsr18/

Keywords

path guiding
realistic rendering

Access to Document

10.2312/sre.20181174

Fingerprint Dive into the research topics of 'Primary Sample Space Path Guiding'. Together they form a unique fingerprint.

Cite this

@inproceedings{425464169d3c47fcb6e29799f3350b9b,

title = "Primary Sample Space Path Guiding",

abstract = "We present a scheme for unbiased path guiding. Different from existing methods that focus on constructing structures in spatial-directional domain, we work in primary sample space. We collect records containing a few dimensions of random numbers as well as the luminance that the resulting path contributes. A multiple dimensional structure is built with collected information. After this, random numbers are drawn from this structure and is used to feed the path tracer. Using this scheme, we are able to work completely outside the rendering kernel. We demonstrate that our method is practical and can be efficient. We manage to reduce variance and reduce zero radiance paths by only working in the primary sample space.",

keywords = "path guiding, realistic rendering",

author = "Guo, {Jerry Jinfeng} and Pablo Bauszat and Jacco Bikker and Elmar Eisemann",

year = "2018",

doi = "10.2312/sre.20181174",

language = "English",

pages = "73--82",

editor = "Wenzel Jakob and Toshiya Hachisuka",

booktitle = "Eurographics Symposium on Rendering - Experimental Ideas and Implementations",

publisher = "Eurographics Association",

note = "EGSR 2018 : 29th Eurographics Symposium on Rendering ; Conference date: 02-07-2018 Through 04-07-2018",

url = "http://cg.ivd.kit.edu/egsr18/",

}

Primary Sample Space Path Guiding. / Guo, Jerry Jinfeng; Bauszat, Pablo; Bikker, Jacco; Eisemann, Elmar.

Eurographics Symposium on Rendering - Experimental Ideas and Implementations. ed. / Wenzel Jakob; Toshiya Hachisuka. Eurographics Association, 2018. p. 73-82.

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

TY - GEN

T1 - Primary Sample Space Path Guiding

AU - Guo, Jerry Jinfeng

AU - Bauszat, Pablo

AU - Bikker, Jacco

AU - Eisemann, Elmar

N1 - Conference code: 29

PY - 2018

Y1 - 2018

N2 - We present a scheme for unbiased path guiding. Different from existing methods that focus on constructing structures in spatial-directional domain, we work in primary sample space. We collect records containing a few dimensions of random numbers as well as the luminance that the resulting path contributes. A multiple dimensional structure is built with collected information. After this, random numbers are drawn from this structure and is used to feed the path tracer. Using this scheme, we are able to work completely outside the rendering kernel. We demonstrate that our method is practical and can be efficient. We manage to reduce variance and reduce zero radiance paths by only working in the primary sample space.

AB - We present a scheme for unbiased path guiding. Different from existing methods that focus on constructing structures in spatial-directional domain, we work in primary sample space. We collect records containing a few dimensions of random numbers as well as the luminance that the resulting path contributes. A multiple dimensional structure is built with collected information. After this, random numbers are drawn from this structure and is used to feed the path tracer. Using this scheme, we are able to work completely outside the rendering kernel. We demonstrate that our method is practical and can be efficient. We manage to reduce variance and reduce zero radiance paths by only working in the primary sample space.

KW - path guiding

KW - realistic rendering

U2 - 10.2312/sre.20181174

DO - 10.2312/sre.20181174

M3 - Conference contribution

SP - 73

EP - 82

BT - Eurographics Symposium on Rendering - Experimental Ideas and Implementations

A2 - Jakob, Wenzel

A2 - Hachisuka, Toshiya

PB - Eurographics Association

T2 - EGSR 2018

Y2 - 2 July 2018 through 4 July 2018

ER -