Robust Regression on Image Manifolds for Ordered Label Denoising

Hui Wu; Richard Souvenir

CVPR 2015poster6 citations

Robust Regression on Image Manifolds for Ordered Label Denoising

Abstract

In this paper, we present a computationally efficient and non-parametric method for robust regression on manifolds. We apply our algorithm to the problem of correcting mislabeled examples from image collections with ordered (e.g., real-valued, ordinal) labels. Compared to related methods for robust regression, our method achieves superior denoising accuracy on a variety of data sets, with label corruption levels as high as 80%. For a diverse set of widely-used, large-scale, publicly-available data sets, our approach results in image labels that more accurately describe the associated images.

BibTeX

@inproceedings{cvpr2015_robustregression,
  title = {Robust Regression on Image Manifolds for Ordered Label Denoising},
  author = {Hui Wu and Richard Souvenir},
  booktitle = {CVPR 2015},
  year = {2015}
}