ML Anthology

Generalized Autoencoder for Volumetric Shape Generation

Yanran Guan, Tansin Jahan, Oliver van Kaick
CVPRW 2020 pp. 1082-1088
doi:10.1109/CVPRW50498.2020.00142 /cvprw/2020/guan2020cvprw-generalized/

Abstract

We introduce a 3D generative shape model based on the generalized autoencoder (GAE). GAEs learn a manifold latent space from data relations explicitly provided during training. In our work, we train a GAE for volumetric shape generation from data similarities derived from the Chamfer distance, and with a loss function which is the combination of the traditional autoencoder loss and the GAE loss. We show that this shape model is able to learn more meaningful structures for the latent manifolds of different categories of shapes, and provides better interpolations between shapes when compared to previous approaches such as autoencoders and variational autoencoders.

PDF CVPRW Semantic Scholar

Cite

Text

Guan et al. "Generalized Autoencoder for Volumetric Shape Generation." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2020. doi:10.1109/CVPRW50498.2020.00142

Markdown

[Guan et al. "Generalized Autoencoder for Volumetric Shape Generation." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2020.](https://mlanthology.org/cvprw/2020/guan2020cvprw-generalized/) doi:10.1109/CVPRW50498.2020.00142

BibTeX

@inproceedings{guan2020cvprw-generalized,
  title     = {{Generalized Autoencoder for Volumetric Shape Generation}},
  author    = {Guan, Yanran and Jahan, Tansin and van Kaick, Oliver},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
  year      = {2020},
  pages     = {1082-1088},
  doi       = {10.1109/CVPRW50498.2020.00142},
  url       = {https://mlanthology.org/cvprw/2020/guan2020cvprw-generalized/}
}