Skip to Main content Skip to Navigation
Conference papers

Spectral clustering of linear subspaces for motion segmentation

Fabien Lauer 1, 2 Christoph Schnörr 1
2 ABC - Machine Learning and Computational Biology
LORIA - Laboratoire Lorrain de Recherche en Informatique et ses Applications
Abstract : This paper studies automatic segmentation of multiple motions from tracked feature points through spectral embedding and clustering of linear subspaces. We show that the dimension of the ambient space is crucial for separability, and that low dimensions chosen in prior work are not optimal. We suggest lower and upper bounds together with a data-driven procedure for choosing the optimal ambient dimension. Application of our approach to the Hopkins155 video benchmark database uniformly outperforms a range of state-of-the-art methods both in terms of segmentation accuracy and computational speed.
Complete list of metadatas

Cited literature [18 references]  Display  Hide  Download
Contributor : Fabien Lauer <>
Submitted on : Thursday, June 18, 2009 - 5:54:40 PM
Last modification on : Tuesday, April 24, 2018 - 1:32:54 PM
Long-term archiving on: : Monday, October 15, 2012 - 2:35:39 PM


Files produced by the author(s)


  • HAL Id : hal-00396782, version 1



Fabien Lauer, Christoph Schnörr. Spectral clustering of linear subspaces for motion segmentation. IEEE Int. Conf. on Computer Vision (ICCV'09), Sep 2009, Kyoto, Japan. ⟨hal-00396782⟩



Record views


Files downloads