TY - GEN
T1 - Local label descriptor for example based semantic image labeling
AU - Yang, Yiqing
AU - Li, Zhouyuan
AU - Zhang, Li
AU - Murphy, Christopher J
AU - Ver Hoeve, Jim
AU - Jiang, Hongrui
PY - 2012
Y1 - 2012
N2 - In this paper we introduce the concept of local label descriptor, which is a concatenation of label histograms for each cell in a patch. Local label descriptors alleviate the label patch misalignment issue in combining structured label predictions for semantic image labeling. Given an input image, we solve for a label map whose local label descriptors can be approximated as a sparse convex combination of exemplar label descriptors in the training data, where the sparsity is regularized by the similarity measure between the local feature descriptor of the input image and that of the exemplars in the training data set. Low-level image over-segmentation can be incorporated into our formulation to improve efficiency. Our formulation and algorithm compare favorably with the baseline method on the CamVid and Barcelona datasets.
AB - In this paper we introduce the concept of local label descriptor, which is a concatenation of label histograms for each cell in a patch. Local label descriptors alleviate the label patch misalignment issue in combining structured label predictions for semantic image labeling. Given an input image, we solve for a label map whose local label descriptors can be approximated as a sparse convex combination of exemplar label descriptors in the training data, where the sparsity is regularized by the similarity measure between the local feature descriptor of the input image and that of the exemplars in the training data set. Low-level image over-segmentation can be incorporated into our formulation to improve efficiency. Our formulation and algorithm compare favorably with the baseline method on the CamVid and Barcelona datasets.
UR - http://www.scopus.com/inward/record.url?scp=84867848661&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84867848661&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-33786-4_27
DO - 10.1007/978-3-642-33786-4_27
M3 - Conference contribution
AN - SCOPUS:84867848661
SN - 9783642337857
VL - 7578 LNCS
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 361
EP - 375
BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
T2 - 12th European Conference on Computer Vision, ECCV 2012
Y2 - 7 October 2012 through 13 October 2012
ER -