Visual Scene Representation with Hierarchical Equivariant Sparse Coding

Abstract

We propose a hierarchical neural network architecture for unsupervised learning of equivariant part-whole decompositions of visual scenes. In contrast to the global equivariance of group-equivariant networks, the proposed architecture exhibits equivariance to part-whole transformations throughout the hierarchy, which we term hierarchical equivariance. The model achieves such internal representations via hierarchical Bayesian inference, which gives rise to rich bottom-up, top-down, and lateral information flows, hypothesized to underlie the mechanisms of perceptual inference in visual cortex. We demonstrate these useful properties of the model on a simple dataset of scenes with multiple objects under independent rotations and translations.

Cite

Text

Shewmake et al. "Visual Scene Representation with Hierarchical Equivariant Sparse Coding." NeurIPS 2023 Workshops: NeurReps, 2023.

Markdown

[Shewmake et al. "Visual Scene Representation with Hierarchical Equivariant Sparse Coding." NeurIPS 2023 Workshops: NeurReps, 2023.](https://mlanthology.org/neuripsw/2023/shewmake2023neuripsw-visual/)

BibTeX

@inproceedings{shewmake2023neuripsw-visual,
  title     = {{Visual Scene Representation with Hierarchical Equivariant Sparse Coding}},
  author    = {Shewmake, Christian A and Buracas, Domas and Lillemark, Hansen and Shin, Jinho and Bekkers, Erik J and Miolane, Nina and Olshausen, Bruno},
  booktitle = {NeurIPS 2023 Workshops: NeurReps},
  year      = {2023},
  url       = {https://mlanthology.org/neuripsw/2023/shewmake2023neuripsw-visual/}
}