Online Neural Path Guiding with Normalized Anisotropic Spherical Gaussians

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Huang, Jiawei, Iizuka, Akito, Tanaka, Hajime, Komura, Taku, Kitamura, Yoshifumi
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914692606197760
author Huang, Jiawei
Iizuka, Akito
Tanaka, Hajime
Komura, Taku
Kitamura, Yoshifumi
author_facet Huang, Jiawei
Iizuka, Akito
Tanaka, Hajime
Komura, Taku
Kitamura, Yoshifumi
contents The variance reduction speed of physically-based rendering is heavily affected by the adopted importance sampling technique. In this paper we propose a novel online framework to learn the spatial-varying density model with a single small neural network using stochastic ray samples. To achieve this task, we propose a novel closed-form density model called the normalized anisotropic spherical gaussian mixture, that can express complex irradiance fields with a small number of parameters. Our framework learns the distribution in a progressive manner and does not need any warm-up phases. Due to the compact and expressive representation of our density model, our framework can be implemented entirely on the GPU, allowing it produce high quality images with limited computational resources.
format Preprint
id arxiv_https___arxiv_org_abs_2303_08064
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Online Neural Path Guiding with Normalized Anisotropic Spherical Gaussians
Huang, Jiawei
Iizuka, Akito
Tanaka, Hajime
Komura, Taku
Kitamura, Yoshifumi
Computer Vision and Pattern Recognition
Graphics
I.3
The variance reduction speed of physically-based rendering is heavily affected by the adopted importance sampling technique. In this paper we propose a novel online framework to learn the spatial-varying density model with a single small neural network using stochastic ray samples. To achieve this task, we propose a novel closed-form density model called the normalized anisotropic spherical gaussian mixture, that can express complex irradiance fields with a small number of parameters. Our framework learns the distribution in a progressive manner and does not need any warm-up phases. Due to the compact and expressive representation of our density model, our framework can be implemented entirely on the GPU, allowing it produce high quality images with limited computational resources.
title Online Neural Path Guiding with Normalized Anisotropic Spherical Gaussians
topic Computer Vision and Pattern Recognition
Graphics
I.3
url https://arxiv.org/abs/2303.08064