Real-Time Path-Guiding Based on Parametric Mixture Models

TL;DR

This paper presents a real-time path-guiding method using online learning of parametric mixture models, significantly improving rendering quality and efficiency in path-tracing by reducing variance, flickering, and computation time.

Contribution

It introduces a novel screen-space technique for real-time path guiding based on online learning of parametric mixture models, enabling efficient variance reduction in path-tracing.

Findings

Achieves up to 4x reduction in FLIP metric at 1 spp

Consumes less than 1.5ms on RTX 2070 for 1080p rendering

Reduces path-tracing timings and flickering significantly

Abstract

Path-Guiding algorithms for sampling scattering directions can drastically decrease the variance of Monte Carlo estimators of Light Transport Equation, but their usage was limited to offline rendering because of memory and computational limitations. We introduce a new robust screen-space technique that is based on online learning of parametric mixture models for guiding the real-time path-tracing algorithm. It requires storing of 8 parameters for every pixel, achieves a reduction of FLIP metric up to 4 times with 1 spp rendering. Also, it consumes less than 1.5ms on RTX 2070 for 1080p and reduces path-tracing timings by generating more coherent rays by about 5% on average. Moreover, it leads to significant bias reduction and a lower level of flickering of SVGF output.