Multiclass Spectral Clustering

  • Authors:
  • Stella X. Yu;Jianbo Shi

  • Affiliations:
  • -;-

  • Venue:
  • ICCV '03 Proceedings of the Ninth IEEE International Conference on Computer Vision - Volume 2
  • Year:
  • 2003

Quantified Score

Hi-index 0.00

Visualization

Abstract

We propose a principled account on multiclass spectral clustering.Given a discrete clustering formulation, we first solve a relaxedcontinuous optimization problem by eigen-decomposition. We clarifythe role of eigenvectors as a generator of all optimal solutionsthrough orthonormal transforms. We then solve an optimaldiscretization problem, which seeks a discrete solution closest tothe continuous optima. The discretization is efficiently computedin an iterative fashion using singular value decomposition andnon-maximum suppression. The resulting discrete solutions arenearly global-optimal. Our method is robust to randominitialization and converges faster than other clustering methods.Experiments on real image segmentation are reported.