The past years have witnessed a large number of interesting applications of various soft computing techniques, such as fuzzy logic, neural networks, and evolutionary computation, to intelligent multimedia processing. This carefully edited monograph presents novel applications of soft computing in multimedia processing. It includes contributions by leading experts in their fields addressing important and timely problems in multimedia computing such as content analysis, indexing and retrieval, recognition and compression, or processing and filtering. This book is aimed at researchers, graduate students, and industrial practitioners in the broad areas of multimedia and soft computing.
Revolutionary computation, fuzzy logic, and probabilistic reasoning. As opposed to conventional "hard" computing, these techniques tolerate imprecision and uncertainty, similar to human beings. In the recent years, successful applications of these powerful methods have been published in many disciplines in numerous journals, conferences, as well as the excellent books in this book series on Studies in Fuzziness and Soft Computing.
This volume is dedicated to recent novel applications of soft computing in multimedia processing. The book is composed of 21 chapters written by experts in their respective fields, addressing various important and timely problems in multimedia computing such as content analysis, indexing and retrieval, recognition and compression, processing and filtering, etc.
In the chapter authored by Guan, Muneesawang, Lay, Amin, and Lee, a radial basis function network with Laplacian mixture model is employed to perform image and video retrieval. D. Androutsos, P. Androutsos, Plataniotis, and Venetsanopoulos investigate color image indexing and retrieval within a small-world framework. Wu and Yap develop a framework of fuzzy relevance feedback to model the uncertainty of users' subjective perception in image retrieval.
Incorporating probabilistic support vector machine and active learning, Chua and Feng present a bootstrapping framework for annotating the semantic concepts of large collections of images. Naphade and Smith expose the challenges of using a support vector machine framework to map low-level media features to high-level semantic concepts for the TREC 2002 benchmark corpus. Song, Lin, and Sun present a cross-modality autonomous learning scheme to build visual semantic models from video sequences or images obtained from the Internet. Xiong, Radhakrishnan, Divakaran, and Huang summarize and compare two of their recent frameworks based on hidden Markov model and Gaussian mixture model for detecting and recognizing "highlight" events in sports videos.