The University of Sheffield
2 files

Mixture of Probabilistic Principal Component Analyzers for Shapes from Point Sets

posted on 2017-06-20, 16:43 authored by Ali Gooya, Alejandro FrangiAlejandro Frangi, Jose Pozo Soler
This Matlab class computes a Mixture of Probabilistic Principal Component Analysers from spatial point clouds with no point-to-point correspondences.

If the point clouds represent shapes, the model generates clusters of shape
atlases, computing mean shape and modes of variations in each cluster of shapes. The naming convention of the variables and the details of the model has been described in the following paper:

"Mixture of Probabilistic Principal Component Analyzers for Shapes from Point Sets" DOI 10.1109/TPAMI.2017.2700276, IEEE Transactions on Pattern Analysis and Machine Intelligence (in press)

A test data set, including point sets from 50 vertebrae models has been included for a quick start under the folder 'SamplePointSets'.

Please refer to the included '' file the for further details on how to run the model. There is no dependencies other than Matlab.


Marie-Curie IIF, (Contract Agreement 625745), granted to A. Gooya



  • There is no personal data or any that requires ethical approval


  • The data complies with the institution and funders' policies on access and sharing

Sharing and access restrictions

  • The data can be shared openly

Data description

  • The file formats are open or commonly used

Methodology, headings and units

  • There is a readme.txt file describing the methodology, headings and units