Deep Shape from Polarization

Yunhao Ba, Alex Gilbert, Franklin Wang, Jinfa Yang, Rui Chen, Yiqin Wang, Lei Yan, Boxin Shi, Achuta Kadambi

ECCV 2020

doi:10.1007/978-3-030-58586-0_33 /eccv/2020/ba2020eccv-deep/

Abstract

This paper makes a first attempt to bring the Shape from Polarization (SfP) problem to the realm of deep learning. The previous state-of-the-art methods for SfP have been purely physics-based. We see value in these principled models, and blend these physical models as priors into a neural network architecture. This proposed approach achieves results that exceed the previous state-of-the-art on a challenging dataset we introduce. This dataset consists of polarization images taken over a range of object textures, paints, and lighting conditions. We report that our proposed method achieves the lowest test error on each tested condition in our dataset, showing the value of blending data-driven and physics-driven approaches.

PDF ECCV Semantic Scholar

Cite

Text

Ba et al. "Deep Shape from Polarization." Proceedings of the European Conference on Computer Vision (ECCV), 2020. doi:10.1007/978-3-030-58586-0_33

Markdown

[Ba et al. "Deep Shape from Polarization." Proceedings of the European Conference on Computer Vision (ECCV), 2020.](https://mlanthology.org/eccv/2020/ba2020eccv-deep/) doi:10.1007/978-3-030-58586-0_33

BibTeX

@inproceedings{ba2020eccv-deep,
  title     = {{Deep Shape from Polarization}},
  author    = {Ba, Yunhao and Gilbert, Alex and Wang, Franklin and Yang, Jinfa and Chen, Rui and Wang, Yiqin and Yan, Lei and Shi, Boxin and Kadambi, Achuta},
  booktitle = {Proceedings of the European Conference on Computer Vision (ECCV)},
  year      = {2020},
  doi       = {10.1007/978-3-030-58586-0_33},
  url       = {https://mlanthology.org/eccv/2020/ba2020eccv-deep/}
}