DIB-R++: Learning to Predict Lighting and Material with a Hybrid Differentiable RendererDownload PDF

21 May 2021, 20:43 (edited 24 Jan 2022)NeurIPS 2021 PosterReaders: Everyone
  • Keywords: differentiable rendering, inverse graphics, material estimation, lighting estimation
  • TL;DR: A hybrid differentiable renderer that supports advanced lighting and material effects and can be embedded in deep learning to jointly predict geometry, light and material from a single image with 2D image supervision only.
  • Abstract: We consider the challenging problem of predicting intrinsic object properties from a single image by exploiting differentiable renderers. Many previous learning-based approaches for inverse graphics adopt rasterization-based renderers and assume naive lighting and material models, which often fail to account for non-Lambertian, specular reflections commonly observed in the wild. In this work, we propose DIBR++, a hybrid differentiable renderer which supports these photorealistic effects by combining rasterization and ray-tracing, taking the advantage of their respective strengths---speed and realism. Our renderer incorporates environmental lighting and spatially-varying material models to efficiently approximate light transport, either through direct estimation or via spherical basis functions. Compared to more advanced physics-based differentiable renderers leveraging path tracing, DIBR++ is highly performant due to its compact and expressive shading model, which enables easy integration with learning frameworks for geometry, reflectance and lighting prediction from a single image without requiring any ground-truth. We experimentally demonstrate that our approach achieves superior material and lighting disentanglement on synthetic and real data compared to existing rasterization-based approaches and showcase several artistic applications including material editing and relighting.
  • Supplementary Material: pdf
  • Code Of Conduct: I certify that all co-authors of this work have read and commit to adhering to the NeurIPS Statement on Ethics, Fairness, Inclusivity, and Code of Conduct.
  • Code: https://nv-tlabs.github.io/DIBRPlus/
15 Replies