Self-supervised Learning for Content-based Image Retrieval
Content-based image retrieval is to retrieve similar images
to the query images based on image contents. It involves some fundamental
problems of image understanding. Previous approaches focused on investigating
the similarity measurements or metric of different images. Many heuristic
methods have been proposed. We treat this problem as a learning problem
by learning the similarities from examples. A classifier could be
trained to classify relevant and irrelevant images. However, one of the
difficulties is that only a very small number of query images are available
as supervised data. It is infeasible to learn meaningful concepts from
such a small data set.
Our idea is to integrating the image database in learning,
i.e., learning is based on both supervised query images and unsupervised
database images. The intuition is that the database with a large number
of unlabeled images could characterize a local probability density that
could be helpful for the hybrid learning. Obviously, if an image is not
in the database, why do need to classify this image? On the other hand,
our learning is only interested in the images in the database, i.e., a
local distribution in the data space. Our proposed Discriminant-EM
(D-EM) algorithm is very suitable for this problem. D-EM inserts a discrimination
step in the EM iteration, to find a discriminating subspace to simplify
the probabilistic structure of the data distribution. D-EM has achieved
very promising results.
Publication:
-
Qi Tian, Ying Wu, Jerry Yu and Thomas S. Huang, "Self-Supervised
Learning Based on Discriminative Nonlinear Features and Its application
for Image Retrieval", in Managing Multimedia Semantics, edited by Uma
Srinivasan and Surya Nepal, ICT Centre CSIRO, Australia, by Idea Group,
Inc., 2004.
-
Qi Tian, Ying Wu, Jerry Yu and Thomas S. Huang, "Self-Supervised
Learning Based on Discriminative Nonlinear Features and Its Applications
for Pattern Classification", Pattern Recognition, Vol.38, No.6,
2005. [PDF]
-
Ying Wu and Thomas S. Huang, "Towards Self-Exploring
Discriminating Features for Visual Learning", Journal of Engineering
Application on Artificial Intelligence, Vol.15, pp.139-150,2002.
[PDF]
-
Qi Tian, Jerry Yu, Ying Wu, and Thomas S. Huang,
"Learning Based on Kernel Discriminant-EM Algorithm for Image Classification",
in Proc. IEEE Int'l Conf. on Acoustics, Speech, and Signal Processing
(ICASSP04), Montreal, Canada, May 2004. [PDF]
-
Ying Wu, Qi Tian, Thomas S. Huang, "Discriminant-EM Algorithm
with Application to Image Retrieval", In Proc. IEEE Conf.
on Computer Vision and Pattern Recognition (CVPR'2000), Hilton Head
Island, SC, June, 2000. [PDF]
-
Ying Wu, Qi Tian and Thomas S. Huang, "Integrating Unlabeled
Images for Image Retrieval Based on Relevance Feedback", In Proc.
of the 15th Int'l Conf. on Pattern Recognition (ICPR'2000), Vol.I,
pp.21-24, Barcelona, Spain, Sept., 2000. [PDF]
-
Qi Tian, Ying Wu and Thomas S. Huang, "Incorporate Discriminant
Analysis with EM Algorithm in Image Retrieval",
In Proc. IEEE Int'l
Conf. on Multimedia and Expo (ICME'2000), New York, July, 2000.
[PDF]
-
Qi Tian, Ying Wu and Thomas S. Huang, "Combine User Defined
Region-of-Interest and Spatial Layout for Image Retrieval", In Proc.
IEEE Int'l Conf. on Image Processing (ICIP'2000), pp. 2061-2064,
Vancouver, Canada, Sept., 2000. [PDF]
-
Ying Wu, Thomas S. Huang, "Using Unlabeled Data in Supervised
Learning by Discriminant-EM Algorithm", Neural Information Processing
Systems (NIPS'99) Workshop on Using Unlabeled Data for Supervised Learning,
Colorado, Dec, 1999.
Return to Video Research
Updated 5/2006. Copyright © 2001-2006 Ying Wu