Manifold Relevance Determination

Andreas Damianou; Carl Henrik Ek; Michalis K. Titsias; Neil D. Lawrence

edit

Back to publications

Manifold Relevance Determination

Andreas Damianou, Carl Henrik Ek, Michalis K. Titsias, Neil D. Lawrence

Proceedings of the International Conference in Machine Learning, Morgan Kauffman 29, 2012.

Abstract

In this paper we present a fully Bayesian latent variable model which exploits conditional nonlinear (in)-dependence structures to learn an efficient latent representation. The latent space is factorized to represent shared and private information from multiple views of the data. In contrast to previous approaches, we introduce a relaxation to the discrete segmentation and allow for a “softly” shared latent space. Further, Bayesian techniques allow us to automatically estimate the dimensionality of the latent spaces. The model is capable of capturing structure underlying extremely high dimensional spaces. This is illustrated by modelling unprocessed images with tenths of thousands of pixels. This also allows us to directly generate novel images from the trained model by sampling from the discovered latent spaces. We also demonstrate the model by prediction of human pose in an ambiguous setting. Our Bayesian framework allows us to perform disambiguation in a principled manner by including latent space priors which incorporate the dynamic nature of the data.

Links

Cite this Paper

BibTeX


@InProceedings{Damianou-manifold12,
  title = 	 {Manifold Relevance Determination},
  author = 	 {Damianou, Andreas and Ek, Carl Henrik and Titsias, Michalis K. and Lawrence, Neil D.},
  booktitle = 	 {Proceedings of the International Conference in Machine Learning},
  year = 	 {2012},
  editor = 	 {Langford, John and Pineau, Joelle},
  volume = 	 {29},
  address = 	 {San Francisco, CA},
  publisher =    {Morgan Kauffman},
  pdf = 	 {https://icml.cc/2012/papers/94.pdf},
  url = 	 {/publications/damianou-manifold12.html},
  abstract = 	 {In this paper we present a fully Bayesian latent variable model which exploits conditional nonlinear (in)-dependence structures to learn an efficient latent representation. The latent space is factorized to represent shared and private information from multiple views of the data. In contrast to previous approaches, we introduce a relaxation to the discrete segmentation and allow for a “softly” shared latent space. Further, Bayesian techniques allow us to automatically estimate the dimensionality of the latent spaces. The model is capable of capturing structure underlying extremely high dimensional spaces. This is illustrated by modelling unprocessed images with tenths of thousands of pixels. This also allows us to directly generate novel images from the trained model by sampling from the discovered latent spaces. We also demonstrate the model by prediction of human pose in an ambiguous setting. Our Bayesian framework allows us to perform disambiguation in a principled manner by including latent space priors which incorporate the dynamic nature of the data.}
}

Endnote

%0 Conference Paper
%T Manifold Relevance Determination
%A Andreas Damianou
%A Carl Henrik Ek
%A Michalis K. Titsias
%A Neil D. Lawrence
%B Proceedings of the International Conference in Machine Learning
%D 2012
%E John Langford
%E Joelle Pineau	
%F Damianou-manifold12
%I Morgan Kauffman
%U /publications/damianou-manifold12.html
%V 29
%X In this paper we present a fully Bayesian latent variable model which exploits conditional nonlinear (in)-dependence structures to learn an efficient latent representation. The latent space is factorized to represent shared and private information from multiple views of the data. In contrast to previous approaches, we introduce a relaxation to the discrete segmentation and allow for a “softly” shared latent space. Further, Bayesian techniques allow us to automatically estimate the dimensionality of the latent spaces. The model is capable of capturing structure underlying extremely high dimensional spaces. This is illustrated by modelling unprocessed images with tenths of thousands of pixels. This also allows us to directly generate novel images from the trained model by sampling from the discovered latent spaces. We also demonstrate the model by prediction of human pose in an ambiguous setting. Our Bayesian framework allows us to perform disambiguation in a principled manner by including latent space priors which incorporate the dynamic nature of the data.

RIS


TY  - CPAPER
TI  - Manifold Relevance Determination
AU  - Andreas Damianou
AU  - Carl Henrik Ek
AU  - Michalis K. Titsias
AU  - Neil D. Lawrence
BT  - Proceedings of the International Conference in Machine Learning
DA  - 2012/06/26
ED  - John Langford
ED  - Joelle Pineau	
ID  - Damianou-manifold12
PB  - Morgan Kauffman
VL  - 29
L1  - https://icml.cc/2012/papers/94.pdf
UR  - /publications/damianou-manifold12.html
AB  - In this paper we present a fully Bayesian latent variable model which exploits conditional nonlinear (in)-dependence structures to learn an efficient latent representation. The latent space is factorized to represent shared and private information from multiple views of the data. In contrast to previous approaches, we introduce a relaxation to the discrete segmentation and allow for a “softly” shared latent space. Further, Bayesian techniques allow us to automatically estimate the dimensionality of the latent spaces. The model is capable of capturing structure underlying extremely high dimensional spaces. This is illustrated by modelling unprocessed images with tenths of thousands of pixels. This also allows us to directly generate novel images from the trained model by sampling from the discovered latent spaces. We also demonstrate the model by prediction of human pose in an ambiguous setting. Our Bayesian framework allows us to perform disambiguation in a principled manner by including latent space priors which incorporate the dynamic nature of the data.
ER  -

APA


Damianou, A., Ek, C.H., Titsias, M.K. & Lawrence, N.D.. (2012). Manifold Relevance Determination. Proceedings of the International Conference in Machine Learning 29 Available from /publications/damianou-manifold12.html.