Volterrafaces Face Recognition System

Discriminant Analysis Using Volterra Kernels

This is a demonstration of the Volterrafaces face recognition system. This complete demo includes data file and can be run with various training and testing set sizes.

The algorithm implemented here was published in the following paper, please cite it if you find this code useful:
Ritwik Kumar, Arunava Banerjee and Baba C. Vemuri, Volterrafaces: Discriminant Analysis using Volterra Kernels”, IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2009.
This paper can be found at www.seas.harvard.edu/~rkkumar.

Details of implementation and code structure is explained in the comments. To run, execute the script volterrafacesDemo.m

volterrafacesDemo.m : Demonstrates Volterraface face recognition system
trainVolterraKernels.m : Helper function for training Volterra classifiers
classifyVolterra.m : Helper function for classifying images using Volterra classifiers
splitImg.m : Helper function for splitting images into patches
transformMatrix.m : Helper function for transforming images into matrix A formulation (Linear)
transformMatrixQuadratic.m : Helper function for transforming images into matrix A formulation (Quadratic)
YaleB32n.mat : Data files containing thumbnails of images used in this demonstration

Copyright (c) Ritwik Kumar, Harvard University 2010