394 search results - page 11 / 79 » Multiclass-Multilabel Classification with More Classes than ... |
VISSYM
2004
14 years 11 months ago
2004
In volume visualization, users typically specify transfer functions to classify the data and assign visual attributes to each material class. Higher-dimensional classification mak...
CVPR
2010
IEEE
15 years 6 months ago
2010
IEEE
Hierarchical conditional random fields have been successfully applied to object segmentation. One reason is their ability to incorporate contextual information at different scales....
COLT
1998
Springer
15 years 1 months ago
1998
Springer
Abstract. We describe several improvements to Freund and Schapire's AdaBoost boosting algorithm, particularly in a setting in which hypotheses may assign confidences to each o...
IJCV
2008
14 years 9 months ago
2008
Multi-class image segmentation has made significant advances in recent years through the combination of local and global features. One important type of global feature is that of i...
ICML
2008
IEEE
15 years 10 months ago
2008
IEEE
Identifying the appropriate kernel function/matrix for a given dataset is essential to all kernel-based learning techniques. A variety of kernel learning algorithms have been prop...