Weakly supervised learning on pre-image problem in kernel methods

Wei Shi Zheng, Jian Huang Lai*, Pong Chi YUEN

*Corresponding author for this work

Research output: Chapter in book/report/conference proceedingConference contributionpeer-review

10 Citations (Scopus)

Abstract

This paper presents a novel alternative approach, namely weakly supervised learning (WSL), to learn the pre-image of a feature vector in the feature space induced by a kernel. It is known that the exact pre-image may typically seldom exist, since the input space and the feature space are not isomorphic in general, and an approximate solution is required in past. The proposed WSL, however, would find an appropriate rather than only a purely approximate solution. WSL is able to involve some weakly supervised prior knowledge into the study of pre-image. The prior knowledge is weak and no class label of the sample is required, providing only information of positive class and negative class which should properly depend on applications. The proposed algorithm is demonstrated on kernel principal component analysis (KPCA) with application to illumination normalization and image denoising on faces. Evaluations of the performance of the proposed algorithm show notable improvement as comparing with some well-known existing approaches.

Original languageEnglish
Title of host publicationProceedings - 18th International Conference on Pattern Recognition, ICPR 2006
Pages711-715
Number of pages5
DOIs
Publication statusPublished - 2006
Event18th International Conference on Pattern Recognition, ICPR 2006 - Hong Kong, China
Duration: 20 Aug 200624 Aug 2006

Publication series

NameProceedings - International Conference on Pattern Recognition
Volume2
ISSN (Print)1051-4651

Conference

Conference18th International Conference on Pattern Recognition, ICPR 2006
Country/TerritoryChina
CityHong Kong
Period20/08/0624/08/06

Scopus Subject Areas

  • Computer Vision and Pattern Recognition

Fingerprint

Dive into the research topics of 'Weakly supervised learning on pre-image problem in kernel methods'. Together they form a unique fingerprint.

Cite this