Volumetric Performance Capture from Minimal Camera Viewpoints

Andrew Gilbert, Marco Volino, John Collomosse, Adrian Hilton

ECCV 2018

doi:10.1007/978-3-030-01252-6_35 /eccv/2018/gilbert2018eccv-volumetric/

Abstract

We present a convolutional autoencoder that enables high fidelity volumetric reconstructions of human performance to be captured from multi-view video comprising only a small set of camera views. Our method yields similar end-to-end reconstruction error to that of a probabilistic visual hull computed using significantly more (double or more) viewpoints. We use a deep prior implicitly learned by the autoencoder trained over a dataset of view-ablated multi-view video footage of a wide range of subjects and actions. This opens up the possibility of high-end volumetric performance capture in on-set and prosumer scenarios where time or cost prohibit a high witness camera count.

PDF ECCV Semantic Scholar

Cite

Text

Gilbert et al. "Volumetric Performance Capture from Minimal Camera Viewpoints." Proceedings of the European Conference on Computer Vision (ECCV), 2018. doi:10.1007/978-3-030-01252-6_35

Markdown

[Gilbert et al. "Volumetric Performance Capture from Minimal Camera Viewpoints." Proceedings of the European Conference on Computer Vision (ECCV), 2018.](https://mlanthology.org/eccv/2018/gilbert2018eccv-volumetric/) doi:10.1007/978-3-030-01252-6_35

BibTeX

@inproceedings{gilbert2018eccv-volumetric,
  title     = {{Volumetric Performance Capture from Minimal Camera Viewpoints}},
  author    = {Gilbert, Andrew and Volino, Marco and Collomosse, John and Hilton, Adrian},
  booktitle = {Proceedings of the European Conference on Computer Vision (ECCV)},
  year      = {2018},
  doi       = {10.1007/978-3-030-01252-6_35},
  url       = {https://mlanthology.org/eccv/2018/gilbert2018eccv-volumetric/}
}