Social media has revolutionized the way people share and access information, from work, daily life, to entertainment. With the ubiquitous presence of capturing devices, multimedia has become the major sharing and interacting medium for information access and communication in various social media networks. There is a clear trend in social media information services for adopting multimedia content. For example, Twitter increases its support on pictures and videos, by releasing its 6-second video sharing app, Vine. Tencent's audio chatting tool, WeChat, has attracted more than 300 million users in less than two years, which is said to replace traditional SMS.
On the other hand, the prominence of social media has also witnessed the social trend in multimedia data generation and consumption. For example, Facebook reportedly sees 350 millions photos uploaded each day as of November 2013; 100 hours of video are uploaded to YouTube every minute, resulting in a total of more than 2 billion videos by the end of 2013. According to Alexa statistics, three out of the four fastest-growing social media websites around 2012-2013 focus on multimedia sharing. Moreover, social media gives birth to many new types of multimedia, e.g., image tweet, audio picture, geo-tagged video, etc. This significantly extends the scope and application areas of multimedia.
The multimedia data generated and consumed under social media circumstances is referred to as social multimedia. Three elements are identified, i.e., multimedia content, user, and interactions (between both user-user and user-content). Impacted by the participatory nature of WEB 2.0, users actively participate in the generation as well as consumption processes. As the hybrid of multimedia and social media, social multimedia enjoys advantages of both direct rich sensory simulation and efficient information access and propagation, thus having great potentials in analysis and utilization.
The emergence of social multimedia has brought challenges as well as opportunities to computing. On one hand, most social multimedia services are user-oriented, making it important to understand user demands from their interactions with the multimedia content. On the other hand, while multimedia content analysis still remains open, the participatory property of social multimedia offers a new solution perspective. Social multimedia computing, a multidisciplinary research and application field, has been developed to understand social multimedia content and connect the social multimedia content with users by exploiting the various social interactions. The potential applications range from information service, communication, entertainment, to healthcare, security, etc.
Thanks to the wide prevalence of social multimedia data and the increasing demands for social multimedia services, there has been a growing number of research on social multimedia computing, evidenced by the volume of papers produced, and many related tracks and special issues in prestigious multimedia conferences and journals. In this tutorial, we will review related work in recent social multimedia computing from two perspectives, i.e. social-sensed multimedia computing, and user-centric social multimedia computing.
Social-sensed multimedia computing. The ultimate goal of multimedia computing is to deliver multimedia content to users according to their information needs (intentions). Multimedia computing can be decomposed into various stages: multimedia compression (for storage), multimedia communication (for delivery), and multimedia content analysis (for intelligence). Among these, multimedia compression and communication are comparatively well established. Since the end of the last century, multimedia analysis has become mainstream in the multimedia community, and related technologies have advanced significantly. However, how to bridge the multimedia content with end users, the lastmile technology for multimedia services, is rarely researched. This negligence directly causes an obvious Intention Gap between multimedia data and the real information needs of users, which has become a bottle-neck in advancing intelligent multimedia computing technologies for use in real applications.
Users' (both crowds' and individuals') intention-related information (including long-term interests, instantaneous intentions, and emotions, etc), their behavior patterns, and ultimately, the common principles of user-multimedia interactions under different contexts can all be sensed from social media, which are summarized by social knowledge on user-multimedia interactions. It is these social knowledge that reflects user needs and establishes a bridge for multimedia data and user needs. How to organically integrate multimedia data, user needs and social knowledge into multimedia computing technology is a critical issue. Here, we propose a new multimedia computing paradigm, social-sensed multimedia computing, to glue together all the recent works that bring social media, a valuable source of sensing user needs and social knowledge, into the loop of multimedia computing. It opens a brand new arena for the multimedia research community with a potentially big impact in both academia and industry. Researchers from the multimedia community have made significant progresses along this direction. The first part of this tutorial aims at: 1) reviewing and summarizing recent high-quality research works on social-sensed multimedia computing, including basic technologies and applicable systems, and 2) presenting insight into the challenges and future directions in this emerging and promising area.
User-centric social multimedia computing. Social multimedia computing is very different from traditional and web multimedia computing. Web multimedia computing is heavily related to the WEB1.0 environment, which is dominated by broadcast media developed by professional designers for passive users. In traditional multimedia computing, the analysis focus is the multimedia content, and the goal is content understanding and application, e.g., media content analysis, semantic classification and annotation, structured media authoring. On the contrary, social multimedia computing has an obvious user-centric characteristic: (1) User is the basic data collection unit. Viewing each user as a data sensor, social multimedia is constituted by what users see, listen, think, feel and speak. Moreover, user bridges multimedia network and social network, contributing to most of the social interactions in social multimedia. (2) User is the ultimate information service target. As discussed above, social multimedia services are user-oriented and have a customized trend. Understanding the personalized demands is critical to most social multimedia computing problems.
We present the user-centric research paradigm for social multimedia computing, and organize the related work into two basic tasks of ``From users, and For users'': (1) From users: user-perceptive multimedia content analysis. Users' social interactions capture what they perceive the multimedia content and can be exploited towards multimedia content analysis. For example, user-contributed picture tags indicate user-perceived visual semantics, user browsing behaviors, such as pause, fast-forward, indicate video structure information. (2) For users: user modeling and customized multimedia applications. Online activities reveals important clues of user background information and are utilized for user modeling of demographic facts, personal interests, social network status, mobility patterns, consuming patters, emotional orientation, etc. Given results from user-perceptive multimedia content analysis and user modeling, customized multimedia services are developed to satisfy personalized needs.
The target audiences of the tutorial include both graduate students and senior researchers working in the field of multimedia and/or social computing, as well as industry practitioners who are working in the field of image/video processing, social applications, search engine development, developers of video/image sharing social platforms, advertising and recommendation systems, and so on.
This tutorial will focus on introducing the important general concepts and themes of this timely topic, and not go deep into the algorithm details. The only prior knowledge expected for the audience is a familiarity with basic machine learning models, e.g., matrix factorization, collaborative filtering, probabilistic density estimation. We expect the attendees will gain (1) a good understanding of the concept of social multimedia and the importance of social multimedia computing; (2) a comprehensive overview of recent high-quality research work organized on social-sensed multimedia computing and user-centric social multimedia computing; and (3) the insight into the challenges and future directions in this emerging and promising area.
Peng Cui is now an Assistant Professor in Tsinghua University, China. He received his PhD degree from Tsinghua University in 2010. He is an active researcher dedicated to novel algorithms and systems in social multimedia computing, and he is keen to promote the convergence of social media data mining and multimedia computing technologies. Dr. Cui has strong backgrounds in both data mining and multimedia communities. He has published more than 30 papers in prestigious conferences and journals in data mining and multimedia, including ACM MM, SIGKDD, SIGIR, AAAI, IEEE TMM, IEEE TKDE, IEEE TIP etc. His recent research won the ACM MM12 Grand Challenge Multimodal Award, and MMM13 Best Paper Award. He is the Area Chair of ACM MM 2014, ICASSP 2013, Associate Editor of Frontier of Computer Science journal, Guest Editor of Information Retrieval journal, and co-organized several special sessions and workshops on social multimedia in ICMR, ICME, ACM MM and WSDM.
Wenwu Zhu is with Computer Science Department of Tsinghua University as Professor of “1000 People Plan” of China. Prior to his current post, he was a Senior Researcher and Research Manager at Microsoft Research Asia. He was the Chief Scientist and the Director at Intel Research China from 2004 to 2008. He worked at Bell Labs New Jersey as Member of Technical Staff during 1996-1999. Wenwu Zhu is an IEEE Fellow, SPIE Fellow and ACM Distinguished Scientist. He has published over 200 referred papers in the areas of multimedia computing, communications and networking. He is inventor or co-inventor of over 40 patents. His current research interests are in the area of multimedia cloud computing, social media computing, multimedia big data, and multimedia communications and networking. He served(s) on various editorial boards, such as Guest Editor for the Proceedings of the IEEE, IEEE T-CSVT, and IEEE JSAC; Associate Editor for IEEE Transactions on Mobile Computing, IEEE Transactions on Multimedia, and IEEE Transactions on Circuits and Systems for Video Technology; Leading Editor of the Area “Computer Networks and Distributed Computing” of Journal of Computer Science and Technology. He received the Best Paper Award in ACM Multimedia 2012, the Best Paper Award in IEEE Transactions on Circuits and Systems for Video Technology in 2001, and the other 3 international Best Paper Awards. He was the Chair of Visual Signal Processing and Communication Technical Committee of IEEE Circuits and Systems Society (2006-2008), and served in the Steering Committee of IEEE Transactions on Mobile Computing (2007-2010). He currently serves as the Chair of Beijing Chapter at IEEE Circuits and Systems Society and advisory board of International Journal of Handheld Computing Research. He served as TPC Co-Chair of IEEE ISCAS 2013 and serves as TPC Co-Chair for ACM Multimedia 2014.
Jitao Sang is assistant professor in National Laboratory of Pattern Recognition at Chinese Academy of Sciences (CAS). He graduated with the highest honor for CAS PhD students, the special prize of CAS president scholarship. His research interest is in social multimedia computing, where the recent research in user-centric social multimedia computing has attracted increasing attentions, with award-winning publications in the prestigious conferences (best paper finalist in MM2012 and MM2013, best student paper in MMM2013). So far, he has authored one book, filed three patents, co-authored more than 40 peer-referenced papers in multimedia-related journals and conferences. He is program co-chair in PCM 2015, ICIMCS 2015, publicity chair in MMM 2015, publication chair in ICIMCS 2013, 2014, special session organizer in ICME2015, MMM2013, ICIMCS 2013, and program committee member in many conferences (MM2013, MM2014, CIKM2014, etc.). He is guest editor in many journals such as MMSJ and MTA. He is keynote speaker at Social Media 2013, and tutorial speaker at MM 2014, MMM 2015, ICMR 2015 and ICME 2015.