Abstract | ||
---|---|---|
Large-scale recognition problems with thousands of classes pose a particular challenge because applying the classifier requires more computation as the number of classes grows. The label tree model integrates classification with the traversal of the tree so that complexity grows logarithmically. In this paper, we show how the parameters of the label tree can be found using maximum likelihood estimation. This new probabilistic learning technique produces a label tree with significantly improved recognition accuracy. |
Year | DOI | Venue |
---|---|---|
2013 | 10.1109/CVPR.2013.114 | CVPR |
Keywords | Field | DocType |
large-scale recognition problem,trees (mathematics),large scale image classification,large scale image,maximum likelihood estimation,particular challenge,probabilistic learning technique,probabilistic label tree model,improved recognition accuracy,image recognition,new probabilistic,label tree,large-scale recognition problems,image classification,large-scale recognition,probabilistic label trees,label tree model,probability,mathematical model,probabilistic logic,accuracy,computational modeling,vectors | Tree traversal,Pattern recognition,Computer science,Decision tree model,Maximum likelihood,Artificial intelligence,Probabilistic logic,Classifier (linguistics),Contextual image classification,Machine learning,Computation | Conference |
Volume | Issue | ISSN |
2013 | 1 | 1063-6919 |
Citations | PageRank | References |
14 | 0.52 | 16 |
Authors | ||
5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Baoyuan Liu | 1 | 132 | 5.64 |
Fereshteh Sadeghi | 2 | 59 | 1.90 |
Marshall F. Tappen | 3 | 1901 | 89.34 |
Ohad Shamir | 4 | 1627 | 119.03 |
Ce Liu | 5 | 3347 | 188.04 |