Feature Generation for Adaptive Gradient-Domain Path Tracing

Jonghee Back1, Sung-Eui Yoon2, Bochang Moon1,

GIST1, KAIST2,

Computer Graphics Forum (Proceedings of Pacific Graphics 2018)

Received Best Paper Honorable Mention award from Pacific Graphics 2018

Figure: Given gradient-domain path tracing (G-PT), we visualize the noisy input color (a) and gradients, (b) and (c), and our feature image (d). Our feature (d) is fed into a state-of-the-art denoising method, adaptive polynomial rendering (APR), and this denoiser greatly reduces a numerical error, i.e., relative mean squared error (relMSE), thanks to our high-quality feature image.

Abstract

In this paper, we propose a new technique to incorporate recent adaptive rendering approaches built upon local regression theory into a gradient-domain path tracing framework, in order to achieve high-quality rendering results. Our method aims to reduce random artifacts introduced by random sampling on image colors and gradients. Our high-level approach is to identify a feature image from noisy gradients, and pass the image to an existing local regression based adaptive method so that adaptive sampling and reconstruction using our feature can boost the performance of gradient-domain rendering. To fulfill our idea, we derive an ideal feature in the form of image gradients and propose an estimation process for the ideal feature in the presence of noise in image gradients. We demonstrate that our integrated adaptive solution leads to performance improvement for a gradient-domain path tracer, by seamlessly incorporating recent adaptive sampling and reconstruction strategies through our estimated feature.

Video (available to download in the last section)

Contents