Towards Conceptual Compression

Karol Gregor; Frederic Besse; Danilo Jimenez Rezende; Ivo Danihelka; Daan Wierstra

Towards Conceptual Compression

Machine Learning 2016-05-02 v1 Computer Vision and Pattern Recognition Machine Learning

Authors: Karol Gregor , Frederic Besse , Danilo Jimenez Rezende , Ivo Danihelka , Daan Wierstra

Abstract

We introduce a simple recurrent variational auto-encoder architecture that significantly improves image modeling. The system represents the state-of-the-art in latent variable models for both the ImageNet and Omniglot datasets. We show that it naturally separates global conceptual information from lower level details, thus addressing one of the fundamentally desired properties of unsupervised learning. Furthermore, the possibility of restricting ourselves to storing only global information about an image allows us to achieve high quality 'conceptual compression'.

Keywords

variational autoencoders deep learning multi-task learning

Cite

@article{arxiv.1604.08772,
  title  = {Towards Conceptual Compression},
  author = {Karol Gregor and Frederic Besse and Danilo Jimenez Rezende and Ivo Danihelka and Daan Wierstra},
  journal= {arXiv preprint arXiv:1604.08772},
  year   = {2016}
}

Comments

14 pages, 13 figures

Towards Conceptual Compression

Abstract

Keywords

Cite

Comments

Related papers