Inducing Interpretable Representations with Variational Autoencoders

By N. Siddharth, Brooks Paige, Alban Desmaison, Jan-Willem van de Meent, Frank D. Wood, Noah D. Goodman, Pushmeet Kohli, Philip H. S. Torr. 2016

7 citations recorded by the source corpus.

We develop a framework for incorporating structured graphical models in the \emph{encoders} of variational autoencoders (VAEs) that allows us to induce interpretable representations through approximate variational inference. This allows us to both perform reasoning (e.g. classification) under the structural constraints of a given graphical model, and use deep generative models to deal with messy, high-dimensional domains where it is often difficult to model all the variation. Learning in this framework is carried out end-to-end with a variational objective, applying to both unsupervised and semi-supervised schemes.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Reconciling deep learning with symbolic artificial intelligence: representing objects and relations (2019, background): “Current Opinion in Behavioral Sciences 2019, 29:17–23 One way to address this issue is to manipulate the loss function being minimised by the learning algorithm so that it favours disentangled representations [20,21 ,22,23] (Figure 1b bottom).”
Black Box Explanation by Learning Image Exemplars in the Latent Feature Space (2019, methodology): “ons of changing some aspect of the AAE: (i) the latent dimensions k, (ii) the rigidity of the discriminator in admitting latent instances,(iii) the type of autoencoders (e.g. variational autoencoders [26]). Second, we would like to extend abele to make it work on tabular data and on text. Third, we ”