license: apache-2.0

task_categories:

• image-to-image

tags:

• datasets
• image-translation
• style-transfer
• material-transfer

pretty_name: MatRender

size_categories:

• 10B<n<100B

Dataset Description

This dataset contains 32 objects, 50 materials, and 10 camera angles.
All images are 512×512 pixels.

• Training set – 8 objects × 10 angles × 50 materials
• Validation set – 12 objects × 10 angles × 50 materials
• Test set – 12 objects × 10 angles × 50 materials

The 3D objects were sourced from free3d.com.

Acknowledgments

This dataset builds upon the rendering pipeline used in:

@article{deschaintre2018single,
  title={Single-image svbrdf capture with a rendering-aware deep network},
  author={Deschaintre, Valentin and Aittala, Miika and Durand, Fredo and Drettakis, George and Bousseau, Adrien},
  journal={ACM Transactions on Graphics (ToG)},
  volume={37},
  number={4},
  pages={1–15},
  year={2018},
  publisher={ACM New York, NY, USA}
}