With the dramatic growth of digital video over the Internet, and rapid development of network and multimedia techniques, making effective use of these vast amounts of data has become a big challenge. This book describes the latest concepts, algorithms, models and applications of semantic video analysis that enable content-based intelligent interpretation and management of video content. It covers seven main research topics including: saliency computation in images and videos; model-based shot boundary detection; extraction of semantic contents and events for video highlights indexing and annotation; activity-driven video summarization; content-based image retrieval using relevance feedback; video retrieval and clustering using fMRI-derived features; and the relatively unexplored area of stereo video generation and retrieval for 3D-TV. Researchers and graduate students, in academia and industry, interested in visual computing, multimedia and computer vision will find this book useful. It will also be of relevance to professionals and practitioners, from fields such as media, broadcasting, museums, film production and online entertainment, who are interested in improving media asset management and content delivery.