Teaser image of "Neural Acceleration of Scattering-Aware Color 3D Printing"
We propose a neural scattering compensation for 3D color printing. Comparing to a method which uses noise-free Monte Carlo simulation our technique achieves 300× speedup in the above case while providing the same quality.

Neural Acceleration of Scattering-Aware Color 3D Printing

Abstract

With the wider availability of full-color 3D printers, color-accurate 3D-print preparation has received increased attention. A key challenge lies in the inherent translucency of commonly used print materials that blurs out details of the color texture. Previous work tries to compensate for these scattering effects through strategic assignment of colored primary materials to printer voxels. To date, the highest-quality approach uses iterative optimization that relies on computationally expensive Monte Carlo light transport simulation to predict the surface appearance from subsurface scattering within a given print material distribution; that optimization, however, takes in the order of days on a single machine. In our work, we dramatically speed up the process by replacing the light transport simulation with a data-driven approach. Leveraging a deep neural network to predict the scattering within a highly heterogeneous medium, our method performs around two orders of magnitude faster than Monte Carlo rendering while yielding optimization results of similar quality level. The network is based on an established method from atmospheric cloud rendering, adapted to our domain and extended by a physically motivated weight sharing scheme that substantially reduces the network size. We analyze its performance in an end-to-end print preparation pipeline and compare quality and runtime to alternative approaches, and demonstrate its generalization to unseen geometry and material values. This for the first time enables full heterogenous material optimization for 3D-print preparation within time frames in the order of the actual printing time.
Play Video
Play Video

What this paper means for Computer Graphics?

This paper is at the overlap between two subfields in Computer Graphics: Color 3D printing using Material Jetting and light transport simulation (or rendering). It proposes to use a model obtained via machine learning to predict the subsurface scattering effects that occur within the 3D printouts. Using this prediction one can compensate for some of the unwanted effects by re-arranging the materials within the object using a virtual optimization.
The model itself is capable of predicting the surface appearance of fully-heterogeneous media (including spatially-varying density and albedo, constant phase function (g=0.4)).

Discussion Page

Eurographics 2021 introduced a page on Reddit where the discussion about the paper can be continued after the conference. 

Press Coverage

We published a press release on the faculty’s webpage, which got picked up by 3dprintingindustry.com3dprintingmedia.network3druck.com (German), 3d-grenzenlos.de (German) and sciencemag.cz (Czech).

BibTex Citation