Subspace Segmentation: Theory and Algorithms

Abstract

Traditional high-dimensional data clustering techniques typically assume that data lives in a single high-dimensional space. However, in many cases, data actually comes from a union of low dimensional subspaces (or manifolds) and some important problems may be cast as a subspace segmentation problem, e.g., motion segmentation, facial expression recognition, recontruction of signals with finite rate of innovation, and compressive sampling modeling. This area of research has recently attracted high interest from computer science, engineering, and applied mathematics. This project will complement and extend theory and techniques from subspace clustering, manifold approximations, and sampling theory. The proposed research will generate interactions between certain areas of mathematics and computer science, such as non-linear approximation, optimization, probability theory, and algorithms. The research goal of this project is to develop mathematical theory and efficient algorithms for subspace segmentation. The PI (Dr. Ali Sekmen) holds two PhD degrees in Electrical Engineering and Mathematics and he has done extensive work on subspace segmentation and its applications to computer vision, signal processing, and bioinformatics. This research will extend our existing theories

Document Details

Document Type
DoD Grant Award
Publication Date
Feb 12, 2016
Source ID
W911NF1510495

Entities

People

  • Ali Sekman

Organizations

  • Army Contracting Command
  • Office of the Secretary of Defense
  • Tennessee State University

Tags

Fields of Study

  • Computer science
  • Engineering

Readers

  • Graph Algorithms and Convex Optimization.
  • Neural Network Machine Learning.
  • Research Science/Academic Research

Technology Areas

  • AI & ML
  • AI & ML - Machine Learning Algorithms
  • Space