DC Field | Value | Language |
---|---|---|
dc.contributor.author | Kim, Tae-Kyun | ko |
dc.contributor.author | Budvytis, Ignas | ko |
dc.contributor.author | Cipolla, Roberto | ko |
dc.date.accessioned | 2021-06-17T06:50:30Z | - |
dc.date.available | 2021-06-17T06:50:30Z | - |
dc.date.created | 2021-06-17 | - |
dc.date.issued | 2012-11 | - |
dc.identifier.citation | INTERNATIONAL JOURNAL OF COMPUTER VISION, v.100, no.2, pp.203 - 215 | - |
dc.identifier.issn | 0920-5691 | - |
dc.identifier.uri | http://hdl.handle.net/10203/285979 | - |
dc.description.abstract | This paper presents a novel way to speed up the evaluation time of a boosting classifier. We make a shallow (flat) network deep (hierarchical) by growing a tree from decision regions of a given boosting classifier. The tree provides many short paths for speeding up while preserving the reasonably smooth decision regions of the boosting classifier for good generalisation. For converting a boosting classifier into a decision tree, we formulate a Boolean optimisation problem, which has been previously studied for circuit design but limited to a small number of binary variables. In this work, a novel optimisation method is proposed for, firstly, several tens of variables i.e. weak-learners of a boosting classifier, and then any larger number of weak-learners by using a two-stage cascade. Experiments on the synthetic and face image data sets show that the obtained tree achieves a significant speed up both over a standard boosting classifier and the Fast-exit-a previously described method for speeding-up boosting classification, at the same accuracy. The proposed method as a general meta-algorithm is also useful for a boosting cascade, where it speeds up individual stage classifiers by different gains. The proposed method is further demonstrated for fast-moving object tracking and segmentation problems. | - |
dc.language | English | - |
dc.publisher | SPRINGER | - |
dc.title | Making a Shallow Network Deep: Conversion of a Boosting Classifier into a Decision Tree by Boolean Optimisation | - |
dc.type | Article | - |
dc.identifier.wosid | 000308364500007 | - |
dc.identifier.scopusid | 2-s2.0-84867101918 | - |
dc.type.rims | ART | - |
dc.citation.volume | 100 | - |
dc.citation.issue | 2 | - |
dc.citation.beginningpage | 203 | - |
dc.citation.endingpage | 215 | - |
dc.citation.publicationname | INTERNATIONAL JOURNAL OF COMPUTER VISION | - |
dc.identifier.doi | 10.1007/s11263-011-0461-z | - |
dc.contributor.localauthor | Kim, Tae-Kyun | - |
dc.contributor.nonIdAuthor | Budvytis, Ignas | - |
dc.contributor.nonIdAuthor | Cipolla, Roberto | - |
dc.description.isOpenAccess | N | - |
dc.type.journalArticle | Article | - |
dc.subject.keywordAuthor | Boosting | - |
dc.subject.keywordAuthor | Decision tree | - |
dc.subject.keywordAuthor | Decision regions | - |
dc.subject.keywordAuthor | Boolean optimisation | - |
dc.subject.keywordAuthor | Boosting cascade | - |
dc.subject.keywordAuthor | Face detection | - |
dc.subject.keywordAuthor | Tracking | - |
dc.subject.keywordAuthor | Segmentation | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.