SurfaceNet: An End-to-End 3D Neural Network for Multiview Stereopsis

Mengqi Ji, Juergen Gall, Haitian Zheng, Yebin Liu, Lu Fang

ICCV 2017

doi:10.1109/ICCV.2017.253 /iccv/2017/ji2017iccv-surfacenet/

Abstract

This paper proposes an end-to-end learning framework for multiview stereopsis. We term the network SurfaceNet. It takes a set of images and their corresponding camera parameters as input and directly infers the 3D model. The key advantage of the framework is that both photo-consistency as well geometric relations of the surface structure can be directly learned for the purpose of multiview stereopsis in an end-to-end fashion. SurfaceNet is a fully 3D convolutional network which is achieved by encoding the camera parameters together with the images in a 3D voxel representation. We evaluate SurfaceNet on the large-scale DTU benchmark.

PDF ICCV Semantic Scholar

Cite

Text

Ji et al. "SurfaceNet: An End-to-End 3D Neural Network for Multiview Stereopsis." International Conference on Computer Vision, 2017. doi:10.1109/ICCV.2017.253

Markdown

[Ji et al. "SurfaceNet: An End-to-End 3D Neural Network for Multiview Stereopsis." International Conference on Computer Vision, 2017.](https://mlanthology.org/iccv/2017/ji2017iccv-surfacenet/) doi:10.1109/ICCV.2017.253

BibTeX

@inproceedings{ji2017iccv-surfacenet,
  title     = {{SurfaceNet: An End-to-End 3D Neural Network for Multiview Stereopsis}},
  author    = {Ji, Mengqi and Gall, Juergen and Zheng, Haitian and Liu, Yebin and Fang, Lu},
  booktitle = {International Conference on Computer Vision},
  year      = {2017},
  doi       = {10.1109/ICCV.2017.253},
  url       = {https://mlanthology.org/iccv/2017/ji2017iccv-surfacenet/}
}