PatchMatchGraph: Building a Graph of Dense Patch Correspondences for Label Transfer
Abstract
We address the problem of semantic segmentation, or multi-class pixel labeling, by constructing a graph of dense overlapping patch correspondences across large image sets. We then transfer annotations from labeled images to unlabeled images using the established patch correspondences. Unlike previous approaches to non-parametric label transfer our approach does not require an initial image retrieval step. Moreover, we operate on a graph for computing mappings between images, which avoids the need for exhaustive pairwise comparisons. Consequently, we can leverage offline computation to enhance performance at test time. We conduct extensive experiments to analyze different variants of our graph construction algorithm and evaluate multi-class pixel labeling performance on several challenging datasets.
Cite
Text
Gould and Zhang. "PatchMatchGraph: Building a Graph of Dense Patch Correspondences for Label Transfer." European Conference on Computer Vision, 2012. doi:10.1007/978-3-642-33715-4_32Markdown
[Gould and Zhang. "PatchMatchGraph: Building a Graph of Dense Patch Correspondences for Label Transfer." European Conference on Computer Vision, 2012.](https://mlanthology.org/eccv/2012/gould2012eccv-patchmatchgraph/) doi:10.1007/978-3-642-33715-4_32BibTeX
@inproceedings{gould2012eccv-patchmatchgraph,
title = {{PatchMatchGraph: Building a Graph of Dense Patch Correspondences for Label Transfer}},
author = {Gould, Stephen and Zhang, Yuhang},
booktitle = {European Conference on Computer Vision},
year = {2012},
pages = {439-452},
doi = {10.1007/978-3-642-33715-4_32},
url = {https://mlanthology.org/eccv/2012/gould2012eccv-patchmatchgraph/}
}