Introduction to the special issue on intelligent multimedia systems and technology

We have witnessed an explosion of multimedia data on the Internet, desktop, and mobile devices in recent years, and there has been an increasing demand for intelligent systems and technology to understand, index, manage, search, and consume these data. Machine learning and data mining have proven to be promising approaches in many data-intensive applications, and many efforts have also been dedicated to multimedia data. The objective of this special issue is to bring together the lat-est research in this direction. For this special issue, we seek effective machine-learning and data-mining algorithms, frameworks, systems, and implementations that particularly work on multimedia data (including image, video, and audio, which may also be associated with textual information). The focus is to identify real challenges in intelligent multimedia systems and technology and to investigate practical solutions to the core problems of multimedia applications in both theoretical and practical perspectives. The articles for this special issue will be published in two issues. This issue contains eight articles that are organized into three parts. The first part contains two articles on machine-learning techniques for multimedia content understanding and search. In the first article, Wang and Hua give a survey on active learning for multimedia annotation and retrieval. Five sample selection strategies are discussed and different learning schemes are compared in this article. Shao et al. propose a transductive inference method with Laplacian regularization for image annotation. They derive two specific algorithms to perform the inference optimization in which the convexity, convergence, and complexity of the algorithms are discussed. Part 2 consists of three articles on the topic of social and Web multimedia analyses and applications. Yu et al. present a novel approach to produce accurate suggestions of suitable " social groups " for a user from the user's personal photo collection. Wu et al. improve automatic photo tagging by learning a distance metric from the implicit side information (visual or textual) vastly available on the social web. In the fourth article, Tang et al. exploit the problem of annotating a large-scale image corpus by label propagation on nosily tagged Web images. Part 3 contains three articles that address three different multimedia applications. Tong et al. present a practical system for automatic player trajectory mapping in soccer videos, based on player detection, unsupervised labeling, multi-object tracking , and playfield registration. Liu et al propose a new JPEG steganalysis approach based on feature mining on discrete cosine transform …