Yangqing Jia and Mathieu Salzmann and Trevor Darrell

EECS Department, University of California, Berkeley

Technical Report No. UCB/EECS-2010-99

June 21, 2010

http://www2.eecs.berkeley.edu/Pubs/TechRpts/2010/EECS-2010-99.pdf

Recent approaches to multi-view learning have shown that factorizing the information into parts that are shared across all views and parts that are private to each view could effectively account for the dependencies and independencies between the different input modalities. Unfortunately, these approaches involve minimizing non-convex objective functions. In this paper, we propose an approach to learning such factorized representations inspired by sparse coding techniques. In particular, we show that structured sparsity allows us to address the multi-view learning problem by alternately solving two convex optimization problems. Furthermore, the resulting factorized latent spaces generalize over existing approaches in that they allow having latent dimensions shared between any subset of the views instead of between all the views only. We show that our approach outperforms state-of-the-art methods on the task of human pose estimation.

---

BibTeX citation:

@techreport{Jia:EECS-2010-99,
    Author= {Jia, Yangqing and Salzmann, Mathieu and Darrell, Trevor},
    Title= {Factorized Latent Spaces with Structured Sparsity},
    Year= {2010},
    Month= {Jun},
    Url= {http://www2.eecs.berkeley.edu/Pubs/TechRpts/2010/EECS-2010-99.html},
    Number= {UCB/EECS-2010-99},
    Abstract= {Recent approaches to multi-view learning have shown that factorizing
the information into parts that are shared across all views and parts
that are private to each view could effectively account for the
dependencies and independencies between the different input
modalities. Unfortunately, these approaches involve minimizing
non-convex objective functions. In this paper, we propose an approach
to learning such factorized representations inspired by sparse coding
techniques. In particular, we show that structured sparsity allows us to
address the multi-view learning problem by alternately solving two
convex optimization problems. Furthermore, the resulting factorized
latent spaces generalize over existing approaches in that they allow
having latent dimensions shared between any subset of the views
instead of between all the views only. We show that our approach
outperforms state-of-the-art methods on the task of human pose
estimation.},
}

EndNote citation:

%0 Report
%A Jia, Yangqing 
%A Salzmann, Mathieu 
%A Darrell, Trevor 
%T Factorized Latent Spaces with Structured Sparsity
%I EECS Department, University of California, Berkeley
%D 2010
%8 June 21
%@ UCB/EECS-2010-99
%U http://www2.eecs.berkeley.edu/Pubs/TechRpts/2010/EECS-2010-99.html
%F Jia:EECS-2010-99