
Matlab classes for Hierarchical Models of Scenes, Objects, & Parts
******************************************************************

Variants of these Matlab classes were used to run the experiments 
reported in the following journal publication:
  Describing Visual Scenes Using Transformed Objects and Parts
  E. B. Sudderth, A. Torralba, W. T. Freeman, and A. S. Willsky
  Int. J. Comp. Vision, vol. 77, pp. 291-330, 2008.

The classes in this directory are used for learning parametric models with
finite, known numbers of objects and parts:
@objectpart:  Parametric, latent Dirichlet allocation (LDA) model for sharing 
              parts among multiple object categories.  IJCV Sec. 3.2, 5.3.
@sceneobject: Parametric, contextual model for scenes containing a fixed,
              known set of object categories.  IJCV Sec. 6, 9.

For demonstrations of MCMC learning on toy data, see the test_objectpart.m
and test_sceneobject.m scripts.  To adapt these scripts to your own data,
replace the call to sampledata.m with appropriate calls to insertdata.m.

For examples of how to use these models with realistic image features,
see the "scripts" directory.

