|
Published Articles >> Table of Contents >> Abstract
18th International Conference on Pattern Recognition (ICPR'06) Volume 2
pp. 711-715
Weakly Supervised Learning on Pre-image Problem in Kernel Methods
Wei-Shi Zheng, Sun Yat-sen University, Guangzhou, P. R. China
Jian-Huang Lai, Sun Yat-sen University, Guangzhou, P. R. China
Pong C. Yuen, Hong Kong Baptist University, Hong Kong
Full Article Text:
  
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/ICPR.2006.1187
Send link to a friend
| 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 preimage
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.
|
 Additional Information
|
Citation:
Wei-Shi Zheng, Jian-Huang Lai, Pong C. Yuen,
"Weakly Supervised Learning on Pre-image Problem in Kernel Methods,"
icpr,
pp. 711-715,
18th International Conference on Pattern Recognition (ICPR'06) Volume 2,
2006
|
|
