Call for Paper
ACM Multimedia Systems Journal
Special Issue on Machine Learning Approaches to Multimedia Information Retrieval


Multimedia information retrieval (MIR) deals with the development of
effective and efficient indexing and retrieval techniques for
multimedia data. The recent explosive growth of the use of multimedia
information in many applications requires advanced techniques for
multimedia information retrieval. However, it is well-known that
research in MIR has been thwarted by the growing semantics gap between
the high level semantics expected from human users and the low level
features currently used for indexing and retrieval of multimedia
information in many existing content based retrieval systems. Recent
development in MIR has revealed that machine learning techniques can
provide effective solutions to diminish this gap. Consequently, many
existing and new research issues in MIR can be effectively tackled by
developing novel machine learning techniques. Such techniques can lead
to the development of new tools that would substantially improve the
retrieval efficiency, effectiveness, and accessibility to multimedia
information databases. For example, recent research has demonstrated
that machine learning techniques, with relevance feedback, can provide
substantial performance improvement for MIR.

The objective of this special issue is to address the current
challenges and research topics in using machine learning techniques to
MIR.

Topics of interest include (but are not limited to): 

     * Active learning in MIR
     * Relevance feedback with learning
     * Incremental learning in MIR
     * Reinforcement learning in MIR
     * Selective sampling methods in MIR
     * Semi-supervised learning in MIR
     * Similarity measures with learning
     * Supervised learning in MIR
     * Unsupervised learning in MIR
     * User modeling in MIR

Authors should follow the ACM Multimedia Systems journal manuscript
format described at the journal Website at
http://cairo.cs.uiuc.edu/mmsj.html. Prospective authors should submit
an electronic copy (PDF or Postscript) of their manuscript to the
online submission system of the journal Website, according to the
following timetable:

* Manuscript due: April 15, 2005
* Notification of acceptance: August 1, 2005
* Final manuscript due: October 1, 2005
* Publication date: Spring 2006

The Webpage of this special issue is located at
http://www.cs.binghamton.edu/~zhongfei/cfp_mmsj_special_issue.html

Guest Editors:

Prof. Arif Ghafoor
School of Electrical and Computer Engineering
Purdue University
ghafoor@ecn.purdue.edu
USA

Prof. Zhongfei (Mark) Zhang
Computer Science Department
Watson School
SUNY Binghamton
zhongfei@cs.binghamton.edu
USA

Prof. Michael S. Lew
LIACS Media Lab
Leiden University
mlew@liacs.nl
The Netherlands

Prof. Zhi-Hua Zhou
Computer Science Department
Nanjing University
zhouzh@nju.edu.cn
China