Skip to Main content Skip to Navigation
Conference papers

Learning Invariant Color Features with Sparse Topographic Restricted Boltzmann Machines

Abstract : Our objective is to learn invariant color features directly from data via unsupervised learning. In this paper, we introduce a method to regularize restricted Boltzmann machines during training to obtain features that are sparse and topographically organized. Upon analysis, the features learned are Gabor-like and demonstrate a coding of orientation, spatial position, frequency and color that vary smoothly with the topography of the feature map. There is also differentiation between monochrome and color filters, with some exhibiting color-opponent properties. We also found that the learned represen- tation is more invariant to affine image transformations and changes in illumination color.
Complete list of metadata
Contributor : Hanlin Goh Connect in order to contact the contributor
Submitted on : Monday, July 9, 2012 - 5:29:35 PM
Last modification on : Thursday, April 21, 2022 - 3:36:02 PM
Long-term archiving on: : Wednesday, October 10, 2012 - 4:25:11 AM


Files produced by the author(s)



Hanlin Goh, Kuśmierz Łukasz, Joo-Hwee Lim, Nicolas Thome, Matthieu Cord. Learning Invariant Color Features with Sparse Topographic Restricted Boltzmann Machines. ICIP 2011 - IEEE International Conference on Image Processing, Sep 2011, Brussels, Belgium. pp.1241-1244, ⟨10.1109/ICIP.2011.6115657⟩. ⟨hal-00716053⟩



Record views


Files downloads