NeuMIP: Multi-Resolution Neural Materials

SIGGRAPH 2021


1University of California, San Diego   2Adobe Research


Abstract

We propose NeuMIP, a neural method for representing and rendering a variety of material appearances at different scales. Classical prefiltering (mipmapping) methods work well on simple material properties such as diffuse color, but fail to generalize to normals, self-shadowing, fibers or more complex microstructures and reflectances. In this work, we generalize traditional mipmap pyramids to pyramids of neural textures, combined with a fully connected network. We also introduce neural offsets, a novel method which enables rendering materials with intricate parallax effects without any tessellation. This generalizes classical parallax mapping, but is trained without supervision by any explicit heightfield. Neural materials within our system support a 7-dimensional query, including position, incoming and outgoing direction, and the desired filter kernel size. The materials have small storage (on the order of standard mipmapping except with more texture channels), and can be integrated within common Monte-Carlo path tracing systems. We demonstrate our method on a variety of materials, resulting in complex appearance across levels of detail, with accurate parallax, self-shadowing, and other effects.


Download

Paper
Arxiv version
Code (coming soon)
Optix Real-Time Demo (coming soon)

Video



Bibtex

@misc{kuznetsov2021neumip,
      title={NeuMIP: Multi-Resolution Neural Materials}, 
      author={Alexandr Kuznetsov and Krishna Mullia and 
         Zexiang Xu and Miloš Hašan and Ravi Ramamoorthi},
      year={2021},
      eprint={2104.02789},
      archivePrefix={arXiv},
      primaryClass={cs.GR}
}