Multimedia Informatics - Introduction to the Special Theme
by Joachim Köhler
Multimedia informatics is a multidisciplinary research area looking at the creation, processing, transmission and consumption of audio-visual data. It combines expertise in signal processing, pattern recognition, coding technology, networking and protocols, data modelling and user interaction. Although these are all challenging research areas in their own right, multimedia informatics stresses their applied usage, exploitation and adaptation. It covers the whole chain of processing, from content authoring to media indexing of the archived content.
Media informatics provides the basis for many practical applications, such as video conferencing, advanced digital TV and Internet services. It has gained from enormous advances in data-capture and presentation hardware, and in broadband network technologies for real-time transmission of multimedia data services.
Different standardization organizations and funding agencies have recognized the important role of multimedia technologies. The MPEG community (http://www.chiariglione.org/mpeg/) was very successful in standardizing advanced audio and video coding technologies. MPEG-2 coding and transmission technology forms the basis for all digital video broadcasting (DVB) services and has generated huge revenue in the media industry. Advanced MPEG-4 audio and video codecs allow the transmission of audio-visual data over mobile networks. The increased availability of multimedia content has also generated the need to manage and organize audio-visual data. This is covered by the MPEG-7 standard, which contains a metadata description scheme to describe multimedia content for indexing and retrieval applications. The goal of many research groups is to invent automatic methods to extract relevant metadata information, using methods from the area of signal processing and pattern recognition. This is a typical example of how algorithms and methods are applied and exploited for multimedia content processing.
Although several applications of multimedia technology already exist in our daily lives, there are still many advances and improvements to be made. This special issue contains 29 articles on a variety of research projects in this field being undertaken by ERCIM members. The invited article by Lynn Wilcox from FXPAL research describes an advanced system for managing digital photo collections using a face recognition engine to detect and classify persons on digital photos. Other work is the result of national and European projects, including SAVANT, MUSCLE and SIMILAR. The importance of international projects is clear, especially when benchmarking tests must be carried out to evaluate the performance of a multimedia system.
The articles are clustered into three areas, which are organized in the following manner:
Topic 1: Multimedia Indexing and Retrieval
This topic contains most of the submitted articles and includes research work on content-based image retrieval, multimedia search engines and the automatic generation and management of metadata. The goal of several research groups is to increase the semantic knowledge of multimedia resources and to close the semantic gap between high-level features coming from text processing, and low-level features from image and audio processing.
Topic 2: Multimedia Networking
Seven articles cover investigations on multimedia networking issues, and present distributed multimedia systems that are connected and managed with intelligent network components (eg proxies). Work on scalable streaming techniques and transcoding mechanisms for adapting the bandwidth for heterogeneous networks is subject of the articles from Sherwin and Bonuccelli.
Topic 3: Interactive Multimedia Applications
These articles describe research work on content authoring and show several interactive multimedia applications. With the toolkit LimSee2, developed by Romain Deltour, it is possible to create multimedia applications and content using the W3C standard SMIL. Another toolkit called TERESA from the research organization ISTI-CNR allows the development of multimodal user interfaces in multi-device environments. The virtual actor Marilyn (see article by Sepideh Chakaveh from Fraunhofer IMK) is applied as a virtual newscaster for an interactive TV application.
Joachim Köhler, Institute for Media Communication -IMK, Fraunhofer ICT Group, Germany