Generalized Principal Component Analysis (GPCA): an Algebraic Geometric Approach to Subspace Clustering
Rene Vidal
Johns Hopkins University
Monday, February 27, 3:00PM
Burchard 124
Computer Science Department
Stevens Institute of Technology
Abstract
This talk will show that for a wide class of segmentation problems (e.g. mixtures of subspaces, mixtures of fundamental matrices/trifocal tensors, mixtures of linear dynamical models) the "chicken-and-egg" dilemma can be tackled using algebraic geometric techniques. In fact, it is possible to eliminate the data segmentation step algebraically and then use all the data to recover all the models without previously segmenting the data as follows:
Applications of GPCA to image/video segmentation, face clustering, 3-D motion segmentation, dynamic texture segmentation, and heart motion analysis will also be presented.