• PDF / 554,531 Bytes
• 13 Pages / 430 x 660 pts Page_size
• 22 Downloads / 210 Views

DOWNLOAD

REPORT

---

Abstract. In retrieval, indexing and classification of multimedia data an efficient information fusion of the different modalities is essential for the system’s overall performance. Since information fusion, its influence factors and performance improvement boundaries have been lively discussed in the last years in different research communities, we will review their latest findings. They most importantly point out that exploiting the feature’s and modality’s dependencies will yield to maximal performance. In data analysis and fusion tests with annotated image collections this is undermined.

1

Introduction

The multi modal nature of multimedia data creates an essential need for information fusion for its classification, indexing and retrieval. Fusion has also great impact on other tasks such as object recognition, since all objects exist in multi modal spaces. Information fusion has established itself as an independent research area over the last decades, but a general formal theoretical framework to describe information fusion systems is still missing [14]. One reason for this is the vast number of disparate research areas that utilize and describe some form of information fusion in their context of theory. For example, the concept of data or feature fusion, which forms together with classifier and decision fusion the three main divisions of fusion levels, initially occurred in multi-sensor processing. By now several other research fields found its application useful. Besides the more classical data fusion approaches in robotics, image processing and pattern recognition, the information retrieval community discovered some years ago its power in combining multiple information sources [23]. The roots of classifier and decision fusion can be found in the neural network literature, where the idea of combining neural network outputs was published as early as 1965 [10]. Later its application expanded into other fields like econometrics as forecast combining, machine learning as evidence combination and also information retrieval in page rank aggregation [23]. In opposite to the early application areas of data, classifier and decision fusion, researchers were for a long time unclear about which level of information fusion is to be preferred and more generally, how to design an optimal information fusion strategy for multimedia processing systems. N. Boujemaa, M. Detyniecki, and A. N¨ urnberger (Eds.): AMR 2007, LNCS 4918, pp. 147–159, 2008. c Springer-Verlag Berlin Heidelberg 2008 

148

J. Kludas, E. Bruno, and S. Marchand-Maillet

This can be seen in recently published approaches that solve similar tasks and nevertheless use different information fusion levels. Examples using classifier fusion are multimedia retrieval [28], multi-modal object recognition [12], multibiometrics [15] and video retrieval [29]. Concerning data fusion, the applications that can be named are multimedia summarization [1], text and image categorization [7], multi-modal image retrieval [27] and web document retrieval [19]. Other problems of interest are th