Part I. Processing and Retrieval (Dr. Androutsos)

The focus of the tutorial is to understand the current research and trends in accessing and retrieving multimedia data based on their content and semantic meaning. The tutorial will address various ways that features andsemantic meaning are identified and indexed and how multimedia applications and services can use this data to efficiently and effectively access the underlying multimedia information. The great surge in the amount of digitaldata that is available today has greatly increased the need for effectively accessing media by their content and meaning. Data is generated both by consumers and industry very easily and very quickly. However, as this amount increases, the inherent value of this multimedia data decreases since it becomes increasingly difficult to find specific images, video scenes or music. If we cannot find the multimedia content that we want, then whatvalue does more and more multimedia data have? Recent research has focused on methods and techniques toautomatically identify salient features and semantic meaning. All multimedia data, i.e., images, video and audio, have low-level content that can be identified and extracted. This content can then be used to retrieve data or can further be used to build high-level semantic meaning. This allows multimedia data to then be retrieved by cognitive concepts such as storyline, audio genre, specific actor or speaker or even by specific actions ofevents (e.g., action segments of a sports broadcast). In this tutorial we will discuss the multimedia retrieval problem. We will investigate various low-level methods that have been developed and how higher-levelsemantic concepts have become the new trend in creating new and exciting ways to manage multimedia,opening doors to new applications.

Part II. Security and Watermarking (Dr. Krishnan)

In recent years, the multimedia content has been reproduced and distributed in digital form. Today’s technology allows people to copy multimedia content and redistribute it over the Internet at a very low cost. Therefore, need for copyright protection has been aroused for maintaining the rights of the multimedia content (audio, image andvideo). Watermarking is the process of embedding a special data into the host signal for copyright ownership. A watermark added to the host signal should satisfy the following conditions. 1) It should be imperceptible by theuser. 2) It should be robust to any kind of manipulations on the host data such as filtering, compression, noise, re-sampling etc. 3) It should be successfully extracted to prove the ownership. In this tutorial, differentwatermarking schemes and their advantages and disadvantages will be covered. Most of the existing algorithms embed the watermark information in the frequency domain of the signal using the simultaneous masking properties. There are also some watermarking techniques proposed that use the temporal maskingproperties. Techniques to embed the watermark in the joint time and frequency domains will also be discussedin the tutorial. Watermarking can only be applied to new media contents and is not feasible for copyright protection of pre-existing digital materials. Multimedia fingerprinting techniques are gaining popularity for information security applications. A Multimedia fingerprint is a compact representation of the perceptually relevant parts of an audio or video content. A well designed fingerprint can be used to identify a multimedia object, even if it is severely distorted by compression or other signal processing techniques. Some of the applications for fingerprinting include integrity verification, content-based identification, broadcast monitoring,and digital rights management. A Fingerprinting system consists of two major blocks: fingerprint extraction and matching. In Fingerprint extraction, using perceptually relevant features of the signal, a compact representationof the signal is developed. In the later part, fast and efficient search and matching criteria are developed tomatch the fingerprint with the fingerprint database. Various fingerprinting schemes including the joint time and frequency based methods, and their advantages and shortcomings will be discussed in the tutorial.

Part III Transmission and Delivery over Access Networks (Dr. Fernando)

In this portion, we focus on the critical challenges facing access networks or the last mile bottle neck formultimedia delivery. We will look at the current and emerging technologies to address them. The dominantbroadband access technologies today are Digital Subscriber Line (DSL) and cable modems. DSL leads, withglobal subscribers exceeding 100 million, while cable-modem subscribers worldwide will total 55 million in 2005.Data over Cable Service Interface Specifications (DOCSIS) is the ITU standard for multimedia delivery overcable. Although DSL is primarily for data, Voice over DSL (VoDSL) is becoming hot in corporate access networks. Furthermore, emerging broadband technologies such as fiber-to-the-home (FTTH) access aregaining in importance. There were 2.8 million FTTH users by March 2005. 30 million people in Japan will have FTTH by 2010 [1]. With FTTH approach, multimedia signals such as (high definition or conventional) video,digital audio and high-speed (Internet) data signals are simultaneously transmitted over optical fiber all the way to the home. The video signal could be either radio frequency (RF) video or internet protocol (IP) video. The IP video is similar to Internet traffic with additional streaming requirement. However, the RF video signal is analog and sub carrier multiplexing is needed to transmit multiple television channels. Furthermore, the RF video is a proven technology that already dominates existing fiber-coaxial networks and has better quality. We will compare the pros and cons of RF video and video over IP technologies for multimedia delivery. Broadband wireless multimedia access is also on the rise. Emerging WiMAX (IEEE 802.16) is a wireless metropolitan-area network technology that provides up to 50- kilometers of service area, without the need of direct line-of-sight to the base station, and provides data rates up to 75 Mbps. Radio over fiber (ROF) technique is another way for broadband access. For example, Chinese government is adapting ROF technology for its future wireless networks. Furthermore, the Federal Communications Commission (FCC) has authorized the use of spectrum at71-76 GHz, 81-86 GHz, and 92-95 GHz in for wireless multimedia delivery in October 2003 that makes 1-10 Gb/s wireless transmission viable. These mm wave will need ROF technology. Reference: [1] HiromichiShinohara (NTT Corp.), ‘Broadband Access in Japan: Rapidly Growing FTTH Market’ IEEE CommunicationMagazine, September 2005, Vol. 43, No. 9, pp: 72-78.

Dr. Dimitri Androutsos (http://www.ee.ryerson.ca/~dimitri) is an Associate Professor at Ryerson University,Toronto, Canada. He is the author of over 50 peer-reviewed international journal articles, conference papers and book chapters. He is a senior member of the IEEE and an executive committee member of the IEEE Toronto Chapter. He has also been on numerous technical committees for international conferences and is a reviewer for many leading international journals. He has also been a recent guest editor for IEEE’s SignalProcessing Magazine on the topic of Semantic Retrieval of Multimedia. In addition, Dr. Androutsos is a Professional Engineer of Ontario and a member of Sigma Xi. He has been with the Department of Electrical & Computer Engineering at Ryerson University since January 2004 where he teaches and performs research in the areas of image, video and multimedia processing. Before his academic position, he worked in industry both in the United States and Canada on signal and image processing projects ranging from object tracking for the film industry to voice-over-IP system development

Dr. Sri Krishnan (http://www.ee.ryerson.ca/~krishnan) Sridhar (Sri) Krishnan received the B.E. degree inElectronics and Communication Engineering from Anna University, Madras, India, in 1993, and the M.S. andPh.D. degrees in Electrical and Computer Engineering from the University of Calgary, Calgary, Alberta, Canada, in 1996 and 1999 respectively. He joined Ryerson University, Toronto, Canada in 1999 as an Assistant Professor in Electrical and Computer Engineering, and is currently an Associate Professor and Chair of the Department. Sri Krishnan's research interests include biomedical signal processing, multimedia watermarking,and biometrics. Three papers authored with his graduate students won the best paper awards adjudicated by IBM T.J. Watson Research in International Conference on Multimedia and Expo (ICME) 2002 held in Switzerland, IEEE Canada in CCECE 2003 held in Montreal, and Micronet, a Network of Centres AnnualWorkshop in 2004. Sri Krishnan is a registered Professional Engineer (P.Eng.) in the Province of Ontario. He is a Senior Member of IEEE, and is associated with the IEEE Engineering in Medicine and Biology Society, IEEE Signal Processing Society, and the IEEE Circuits and Systems Society. He was the IEEE Student Branch Counselor of Ryerson University from 2001 to 2003. He is the Founding General Chairman of the International Conference for Upcoming Engineers (ICUE) hosted annually by an IEEE student branch in Region 7 (Canada). From October 2003 to October 2005 he served as the Chair of the IEEE Signals and Applications Chapter of Toronto Section. This joint chapter represented six IEEE societies, and in 2005, Sri Krishnan was recognized with a Certificate of Appreciation by the six parent societies for his notable and outstanding contributions toIEEE and the Engineering profession. Since October 2005 he is serving as the Chair of the newly formed IEEE Signal Processing Chapter of the Toronto Section.

Dr. Xavier Fernando (http://www.ee.ryerson.ca/~fernando) is an Associate Professor and Coordinator of the Signal Processing and Communications Research Stream at Ryerson University, Toronto, Canada. He is the author of about forty peer-reviewed journal articles and conference papers. He has one patent for optical fiber based wireless access scheme. He won the best paper award in Canadian Conference on Electrical andComputer Engineering (CCECE - 2001). He has worked for AT&T for three years as an R&D Engineer. He is aSenior Member of IEEE, registered Professional Engineer in Canada. He is the Chair of the IEEECommunications Society, Toronto Chapter. His research focuses on signal processing for cost-effective broadband multimedia delivery via optical, copper and wireless networks. His student projects won both the first and second prize at Opto Canada – the SPIE regional conference in Ottawa in 2002. He has won the research grants and awards from Finnish International Development Agency (FINNIDA), Alberta Informatics Circle of Research Excellence (ICORE), Natural Sciences and Engineering Research Council (NSERC) of Canada,Canadian Foundation of Innovations (CFI) and Ontario Innovations Trust Fund (OIT).