Papers by Jan Novak


Order by: Date  | Author  | Title  | Index of all authors  | Index of Pixar Technical Memos


Denoising with Kernel Prediction and Asymmetric Loss Functions

Thijs Vogels, Fabrice Rouselle, Brian McWilliams, Gerhard Roethlin, Alex Harvill, David Adler, Mark Meyer, Jan Novak
May 2018

We present a modular convolutional architecture for denoising rendered images. We expand on the capabilities of kernel-predicting networks by combining them with a number of task-specific modules, and optimizing the assembly using an asymmetric loss. The source aware encoder - the first module in the assembly - extracts ... more

Paper (PDF)


Kernel-Predicting Convolutional Networks for Denoising Monte Carlo Renderings

Steve Bako, Thijs Vogels, Brian McWilliams, Mark Meyer, Jan Novak, Alex Harvill, Pradeep Sen, Tony DeRose, Fabrice Rousselle
July 2017

Regression-based algorithms have shown to be good at denoising Monte Carlo (MC) renderings by leveraging its inexpensive by-products (e.g., feature buffers). However, when using higher-order models to handle complex cases, these techniques often overfit to noise in the input. For this reason, supervised learning methods have been proposed that ... more

Paper (PDF)

SIGGRAPH 2017