Active rare class discovery and classification using Dirichlet processes

Tom S.F, Haines, Tao Xiang

Research output: Contribution to journalArticle

11 Citations (Scopus)
60 Downloads (Pure)

Abstract

Classification is used to solve countless problems. Many real world computer vision problems, such as visual surveillance, contain uninteresting but common classes alongside interesting but rare classes. The rare classes are often unknown, and need to be discovered whilst training a classifier. Given a data set active learning selects the members within it to be labelled for the purpose of constructing a classifier, optimising the choice to get the best classifier for the least amount of effort. We propose an active learning method for scenarios with unknown, rare classes, where the problems of classification and rare class discovery need to be tackled jointly. By assuming a non-parametric prior on the data the goals of new class discovery and classification refinement are automatically balanced, without any tunable parameters. The ability to work with any specific classifier is maintained, so it may be used with the technique most appropriate for the problem at hand. Results are provided for a large variety of problems, demonstrating superior performance.
Original languageEnglish
Pages (from-to)315-331
Number of pages18
JournalInternational Journal of Computer Vision
Volume106
Issue number3
Early online date23 May 2013
DOIs
Publication statusPublished - 1 Feb 2014

Fingerprint

Classifiers
Computer vision
Problem-Based Learning

Keywords

  • Active learning
  • Rare class discovery
  • Classification

Cite this

Active rare class discovery and classification using Dirichlet processes. / Haines, Tom S.F,; Xiang, Tao.

In: International Journal of Computer Vision, Vol. 106, No. 3, 01.02.2014, p. 315-331.

Research output: Contribution to journalArticle

@article{935d6450f9d34cb7ae76ae13947d5402,
title = "Active rare class discovery and classification using Dirichlet processes",
abstract = "Classification is used to solve countless problems. Many real world computer vision problems, such as visual surveillance, contain uninteresting but common classes alongside interesting but rare classes. The rare classes are often unknown, and need to be discovered whilst training a classifier. Given a data set active learning selects the members within it to be labelled for the purpose of constructing a classifier, optimising the choice to get the best classifier for the least amount of effort. We propose an active learning method for scenarios with unknown, rare classes, where the problems of classification and rare class discovery need to be tackled jointly. By assuming a non-parametric prior on the data the goals of new class discovery and classification refinement are automatically balanced, without any tunable parameters. The ability to work with any specific classifier is maintained, so it may be used with the technique most appropriate for the problem at hand. Results are provided for a large variety of problems, demonstrating superior performance.",
keywords = "Active learning, Rare class discovery, Classification",
author = "Haines, {Tom S.F,} and Tao Xiang",
year = "2014",
month = "2",
day = "1",
doi = "10.1007/s11263-013-0630-3",
language = "English",
volume = "106",
pages = "315--331",
journal = "International Journal of Computer Vision",
issn = "0920-5691",
publisher = "Springer Netherlands",
number = "3",

}

TY - JOUR

T1 - Active rare class discovery and classification using Dirichlet processes

AU - Haines, Tom S.F,

AU - Xiang, Tao

PY - 2014/2/1

Y1 - 2014/2/1

N2 - Classification is used to solve countless problems. Many real world computer vision problems, such as visual surveillance, contain uninteresting but common classes alongside interesting but rare classes. The rare classes are often unknown, and need to be discovered whilst training a classifier. Given a data set active learning selects the members within it to be labelled for the purpose of constructing a classifier, optimising the choice to get the best classifier for the least amount of effort. We propose an active learning method for scenarios with unknown, rare classes, where the problems of classification and rare class discovery need to be tackled jointly. By assuming a non-parametric prior on the data the goals of new class discovery and classification refinement are automatically balanced, without any tunable parameters. The ability to work with any specific classifier is maintained, so it may be used with the technique most appropriate for the problem at hand. Results are provided for a large variety of problems, demonstrating superior performance.

AB - Classification is used to solve countless problems. Many real world computer vision problems, such as visual surveillance, contain uninteresting but common classes alongside interesting but rare classes. The rare classes are often unknown, and need to be discovered whilst training a classifier. Given a data set active learning selects the members within it to be labelled for the purpose of constructing a classifier, optimising the choice to get the best classifier for the least amount of effort. We propose an active learning method for scenarios with unknown, rare classes, where the problems of classification and rare class discovery need to be tackled jointly. By assuming a non-parametric prior on the data the goals of new class discovery and classification refinement are automatically balanced, without any tunable parameters. The ability to work with any specific classifier is maintained, so it may be used with the technique most appropriate for the problem at hand. Results are provided for a large variety of problems, demonstrating superior performance.

KW - Active learning

KW - Rare class discovery

KW - Classification

U2 - 10.1007/s11263-013-0630-3

DO - 10.1007/s11263-013-0630-3

M3 - Article

VL - 106

SP - 315

EP - 331

JO - International Journal of Computer Vision

JF - International Journal of Computer Vision

SN - 0920-5691

IS - 3

ER -