AISTATS 2019poster63 citations

Uncertainty Autoencoders: Learning Compressed Representations via Variational Information Maximization

Aditya Grover, Stefano Ermon

Abstract

Compressed sensing techniques enable efficient acquisition and recovery of sparse, highdimensional data signals via low-dimensional projections. In this work, we propose Uncertainty Autoencoders, a learning framework for unsupervised representation learning inspired by compressed sensing. We treat the low-dimensional projections as noisy latent representations of an autoencoder and directly learn both the acquisition (i.e., encoding) and amortized recovery (i.e., decoding) procedures. Our learning objective optimizes for a tractable variational lower bound to the mutual information between the datapoints and the latent representations. We show how our framework provides a unified treatment to several lines of research in dimensionality reduction, compressed sensing, and generative modeling. Empirically, we demonstrate a 32% improvement on average over competing approaches for the task of statistical compressed sensing of high-dimensional datasets.

BibTeX
@InProceedings{pmlr-v89-grover19a,
  title = 	 {Uncertainty Autoencoders: Learning Compressed Representations via Variational Information Maximization},
  author =       {Grover, Aditya and Ermon, Stefano},
  booktitle = 	 {Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics},
  pages = 	 {2514--2524},
  year = 	 {2019},
  editor = 	 {Chaudhuri, Kamalika and Sugiyama, Masashi},
  volume = 	 {89},
  series = 	 {Proceedings of Machine Learning Research},
  month = 	 {16--18 Apr},
  publisher =    {PMLR},
  pdf = 	 {http://proceedings.mlr.press/v89/grover19a/grover19a.pdf},
  url = 	 {https://proceedings.mlr.press/v89/grover19a.html},
  abstract = 	 {Compressed sensing techniques enable efficient acquisition and recovery of sparse, highdimensional data signals via low-dimensional projections. In this work, we propose Uncertainty Autoencoders, a learning framework for unsupervised representation learning inspired by compressed sensing. We treat the low-dimensional projections as noisy latent representations of an autoencoder and directly learn both the acquisition (i.e., encoding) and amortized recovery (i.e., decoding) procedures. Our learning objective optimizes for a tractable variational lower bound to the mutual information between the datapoints and the latent representations. We show how our framework provides a unified treatment to several lines of research in dimensionality reduction, compressed sensing, and generative modeling. Empirically, we demonstrate a 32% improvement on average over competing approaches for the task of statistical compressed sensing of high-dimensional datasets.}
}
Uncertainty Autoencoders: Learning Compressed Representations via Variational Information Maximization · AISTATS 2019