We address the problem of segmenting 3D scan data into objects or object classes. Our segmentation framework is based on a subclass of Markov Random Fields (MRFs) which support efficient graph-cut inference. The MRF models incorporate a large set of diverse features and enforce the preference that adjacent scan points have the same classification label. We use a recently proposed maximummargin framework to discriminatively train the model from a set of labeled scans; as a result we automatically learn the relative importance of the features for the segmentation task. Performing graph-cut inference in the trained MRF can then be used to segment new scenes very efficiently. We test our approach on three large-scale datasets produced by different kinds of 3D sensors, showing its applicability to both outdoor and indoor environments containing diverse objects.
Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05)
year
2005
pages
169--176
publisher
IEEE Computer Society
volume
2
issn
1063-6919
file
:E\:\Łiteratursammlung\\Proceedings\\Sonstige\\Anguelov et al (2005) - Discriminative learning of markov random fields for segmentation of 3d scan data.pdf:PDF
%0 Conference Paper
%1 Anguelov2005
%A Anguelov, D.
%A Taskar, B.
%A Chatalbashev, V.
%A Koller, D.
%A Gupta, D.
%A Heitz, G.
%A Ng, A.
%B Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05)
%D 2005
%I IEEE Computer Society
%K Learning Laserscanning segmentation 3DPointCloud
%P 169--176
%T Discriminative learning of markov random fields for segmentation of 3d scan data
%V 2
%X We address the problem of segmenting 3D scan data into objects or object classes. Our segmentation framework is based on a subclass of Markov Random Fields (MRFs) which support efficient graph-cut inference. The MRF models incorporate a large set of diverse features and enforce the preference that adjacent scan points have the same classification label. We use a recently proposed maximummargin framework to discriminatively train the model from a set of labeled scans; as a result we automatically learn the relative importance of the features for the segmentation task. Performing graph-cut inference in the trained MRF can then be used to segment new scenes very efficiently. We test our approach on three large-scale datasets produced by different kinds of 3D sensors, showing its applicability to both outdoor and indoor environments containing diverse objects.
@inproceedings{Anguelov2005,
abstract = {We address the problem of segmenting 3D scan data into objects or object classes. Our segmentation framework is based on a subclass of Markov Random Fields (MRFs) which support efficient graph-cut inference. The MRF models incorporate a large set of diverse features and enforce the preference that adjacent scan points have the same classification label. We use a recently proposed maximummargin framework to discriminatively train the model from a set of labeled scans; as a result we automatically learn the relative importance of the features for the segmentation task. Performing graph-cut inference in the trained MRF can then be used to segment new scenes very efficiently. We test our approach on three large-scale datasets produced by different kinds of 3D sensors, showing its applicability to both outdoor and indoor environments containing diverse objects.},
added-at = {2010-12-09T16:35:40.000+0100},
author = {Anguelov, D. and Taskar, B. and Chatalbashev, V. and Koller, D. and Gupta, D. and Heitz, G. and Ng, A.},
biburl = {https://www.bibsonomy.org/bibtex/262aa4351944f54249251815cd4af530f/ipi_jn},
booktitle = {Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05)},
file = {:E\:\\Literatursammlung\\Proceedings\\Sonstige\\Anguelov et al (2005) - Discriminative learning of markov random fields for segmentation of 3d scan data.pdf:PDF},
interhash = {d7703ff1a719fe7460f838d6fa4f0fbb},
intrahash = {62aa4351944f54249251815cd4af530f},
issn = {1063-6919},
keywords = {Learning Laserscanning segmentation 3DPointCloud},
pages = {169--176},
publisher = {IEEE Computer Society},
timestamp = {2010-12-09T16:35:40.000+0100},
title = {Discriminative learning of markov random fields for segmentation of 3d scan data},
volume = 2,
year = 2005
}