Grassmannian Hashing

Grassmannian Hashing (GH) is an approximate nearest subspace search algorithm for solving the problem of retrieving similar subspaces via an approximation scheme called Locality Sensitive Hashing. For the detailed description of the algorithm and its numerical experiments, please refer to the ICCV paper below. For the justification of the algorithm and its underlying theory, please refer to the supplementary file.

Matlab code for Grassmannian Hashing: GH.
Brief description of the algorithm: slides.

Data Sets

The following are links to the data sets used in this work: Extended Yale B dataset, Berkeley segmentation dataset, MultiPie dataset. For use of these data sets, please refer to the original websites for details.

The material presented in this work is partially supported by NSF grants DMS-09-15064 and DMS-09-56072, and ONR N00014-13-1-0492.


Back to Gilad Lerman's webpage

Last Modified Monday April 24, 2017
The views and opinions expressed in this page are strictly those of the page author. The contents of this page have not been reviewed or approved by the University of Minnesota.