2002
DOI: 10.1111/1467-8659.00703
|View full text |Cite
|
Sign up to set email alerts
|

A Simple and Robust Mutation Strategy for the Metropolis Light Transport Algorithm

Abstract: This paper presents a new mutation strategy for the Metropolis light transport algorithm, which works in the unit

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
2
1

Citation Types

0
41
0
6

Year Published

2010
2010
2020
2020

Publication Types

Select...
6
1

Relationship

0
7

Authors

Journals

citations
Cited by 34 publications
(47 citation statements)
references
References 3 publications
0
41
0
6
Order By: Relevance
“…Although capable of dealing with complex situations, the main issue for the basic BDPT is excessive noise. This has been addressed by variety of improvement techniques such as Metropolis Light Transport (MLT) [49] [50], Gradient-Domain Bidirectional Path Tracing (G-BDPT) [51], Energy Redistribution Path Tracing (ERPT) [52], Gradient-domain Metropolis Light Transport (G-MLT) [53], Primary Space Simple Metropolis Light Transport (PSSMLT) [54] and particularly interesting Manifold Exploration Path Tracing (MEPT) [55]. Latter chooses mutation in paths to calculate specular regions and caustics and produce smooth results efficiently.…”
Section: Stochastic (Monte Carlo) Based Methods -Ray Tracingmentioning
confidence: 99%
“…Although capable of dealing with complex situations, the main issue for the basic BDPT is excessive noise. This has been addressed by variety of improvement techniques such as Metropolis Light Transport (MLT) [49] [50], Gradient-Domain Bidirectional Path Tracing (G-BDPT) [51], Energy Redistribution Path Tracing (ERPT) [52], Gradient-domain Metropolis Light Transport (G-MLT) [53], Primary Space Simple Metropolis Light Transport (PSSMLT) [54] and particularly interesting Manifold Exploration Path Tracing (MEPT) [55]. Latter chooses mutation in paths to calculate specular regions and caustics and produce smooth results efficiently.…”
Section: Stochastic (Monte Carlo) Based Methods -Ray Tracingmentioning
confidence: 99%
“…Именно так и было сделано в самой первой работе по MLT [19], где алгоритм Метрополи-са был предложен для BPT и использованы переходы в пространстве путей (path space) как небольшие изменения позиций вершин пути. В дальнейшем в работе [23] алгоритм Метрополиса был применён и к однонаправленной, и к двунаправленной трассировке путей с многократной выборкой по значимости. При этом уже были использованы переходы в так называемом первичном про-странстве путей (primary sample space, PSSMLT) -многомерном единичном кубе, в котором генерировались случайные числа.…”
Section: 2unclassified
“…При этом уже были использованы переходы в так называемом первичном про-странстве путей (primary sample space, PSSMLT) -многомерном единичном кубе, в котором генерировались случайные числа. Причём в работе [23] допол-нительно предлагается смешивать результаты работы алгоритма Метрополиса и обычного Монте-Карло при помощи того же самого механизма многократной выборки по значимости, что улучшает свойства алгоритма в тёмных участках изображения.…”
Section: 2unclassified
“…A variant of MLT, which can be simpler to implement in practice is the Primary Sample Space MLT (PSSMLT) algorithm proposed by Kelemen et al [110]. Instead of directly proposing changes to the path vertices, this method is based on The images were all rendered using the Mitsuba render [94] in five minutes using 64 Amazon EC2 vCPUs.…”
Section: Primary Sample Space Mlt (Pssmlt)mentioning
confidence: 99%
“…where p(x) is the path sampling probability of the underlying path tracer, for more details we refer to [110]. This integral can now be approximated using MCMC by sampling proportional to…”
Section: Primary Sample Space Mlt (Pssmlt)mentioning
confidence: 99%