Sparse decomposition of transformation-invariant signals with continuous basis pursuit

Consider the decomposition of a signal into features that undergo transformations drawn from a continuous family. Current methods discretely sample the transformations and apply sparse recovery methods to the resulting finite dictionary. These methods do not exploit the underlying continuous structu...

Full description

Saved in:
Bibliographic Details
Main Authors: Ekanadham, Chaitanya, Tranchina, Daniel, Simoncelli, Eero P.
Format: Conference Proceeding
Language:English
Subjects:
Accuracy
basis pursuit
Dictionaries
feature decomposition
Interpolation
invariance
Manifolds
Signal to noise ratio
sparsity
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Consider the decomposition of a signal into features that undergo transformations drawn from a continuous family. Current methods discretely sample the transformations and apply sparse recovery methods to the resulting finite dictionary. These methods do not exploit the underlying continuous structure, thereby limiting the ability to produce sparse solutions. Instead, we employ interpolation functions which linearly approximate the manifold of scaled and transformed features. Coefficients are interpreted as interpolation weights, and we formulate a convex optimization problem for obtaining them, enforcing both reconstruction accuracy and sparsity. We compare our method, which we call continuous basis pursuit (CBP) with the standard basis pursuit approach on a sparse deconvolution task. CBP yields substantially sparser solutions without sacrificing accuracy, and does so with a smaller dictionary. We conclude that for signals generated by transformation-invariant processes, a representation that explicitly accommodates the transformation(s) can yield sparser and more interpretable decompositions.
ISSN:1520-6149
2379-190X
DOI:10.1109/ICASSP.2011.5947244