Unsupervised Learning of Geometrical Features from Images by Explicit Group Actions Enforcement
Abstract
In this work we propose an autoencoder architecture capable of automatically learning meaningful geometric features of objects in images, achieving a disentangled representation of 2D objects. It is made of a standard dense autoencoder that captures the deep features identifying the shapes and an additional encoder that extracts geometric latent variables regressed in an unsupervised manner. These are then used to apply a transformation on the output of the deep features decoder. The promising results show that this approach performs better than a non-constrained model having more degrees of freedom.
Cite
Text
Calisto et al. "Unsupervised Learning of Geometrical Features from Images by Explicit Group Actions Enforcement." NeurIPS 2022 Workshops: NeurReps, 2022.Markdown
[Calisto et al. "Unsupervised Learning of Geometrical Features from Images by Explicit Group Actions Enforcement." NeurIPS 2022 Workshops: NeurReps, 2022.](https://mlanthology.org/neuripsw/2022/calisto2022neuripsw-unsupervised/)BibTeX
@inproceedings{calisto2022neuripsw-unsupervised,
title = {{Unsupervised Learning of Geometrical Features from Images by Explicit Group Actions Enforcement}},
author = {Calisto, Francesco and Bottero, Luca and Pagliarino, Valerio},
booktitle = {NeurIPS 2022 Workshops: NeurReps},
year = {2022},
url = {https://mlanthology.org/neuripsw/2022/calisto2022neuripsw-unsupervised/}
}