Compact Matrix Factorization with Dependent Subspaces
Paper in proceeding, 2017

Traditional matrix factorization methods approximate high dimensional data with a low dimensional subspace. This imposes constraints on the matrix elements which allow for estimation of missing entries. A lower rank provides stronger constraints and makes estimation of the missing entries less ambiguous at the cost of measurement fit. In this paper we propose a new factorization model that further constrains the matrix entries. Our approach can be seen as a unification of traditional low-rank matrix factorization and the more recent union-of-subspace approach. It adaptively finds clusters that can be modeled with low dimensional local subspaces and simultaneously uses a global rank constraint to capture the overall scene interactions. For inference we use an energy that penalizes a trade-off between data fit and degrees-of-freedom of the resulting factorization. We show qualitatively and quantitatively that regularizing both local and global dynamics yields significantly improved missing data estimation.

Author

Viktor Larsson

Carl Olsson

Chalmers, Signals and Systems

30th IEEE Conference on Computer Vision and Pattern Recognition

1063-6919 (ISSN)

Vol. 2017-January 4361-4370
978-1-5386-0457-1 (ISBN)

Subject Categories

Computer Engineering

DOI

10.1109/cvpr.2017.464

ISBN

978-1-5386-0457-1

More information

Latest update

7/11/2024