Hierarchical Dictionary Learning for Invariant Classification

Abstract

Sparse representation theory has been increasingly used in the fields of signal processing and machine learning. The standard sparse models are not invariant to spatial transformations such as image rotations, and the representation is very sensitive even under small such distortions. Most studies addressing this problem proposed algorithms which either use transformed data as part of the training set, or are invariant or robust only under minor transformations. In this paper we suggest a framework which extracts invariant sparse features under significant rotations and scalings. The algorithm is based on a hierarchical architecture of dictionary learning for sparse coding in a cortical (log-polar) space. The proposed model is tested in supervised classification applications and proved to be robust under transformed data sets.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 2009
Accession Number
ADA513255

Entities

People

  • Guillermo Sapiro
  • Leah Bar

Organizations

  • University of Minnesota

Tags

Communities of Interest

  • Autonomy

DTIC Thesaurus Topics

  • Algorithms
  • Cartography
  • Classification
  • Coefficients
  • Complex Numbers
  • Compressed Sensing
  • Computer Programming
  • Data Sets
  • Dictionaries
  • Feature Extraction
  • Hierarchies
  • Image Processing
  • Invariance
  • Learning
  • Machine Learning
  • Numbers
  • Pattern Recognition

Fields of Study

  • Computer science

Readers

  • Computer Vision.

Technology Areas

  • AI & ML
  • Space