Video Search and Mining

As cameras become more pervasive in our daily life, vast amounts of video data are generated. The popularity of YouTube and similar websites such as Tudou and Youku provides strong evidence for the increasing role of video in society. One of the main challenges confronting us in the era of information technology is to - fectively rely on the huge and rapidly growing video data accumulating in large multimedia archives. Innovative video processing and analysis techniques will play an increasingly important role in resolving the difficult task of video search and retrieval. A wide range of video-based applications have benefited from - vances in video search and mining including multimedia information mana- ment, human-computer interaction, security and surveillance, copyright prot- tion, and personal entertainment, to name a few. This book provides an overview of emerging new approaches to video search and mining based on promising methods being developed in the computer vision and image analysis community. Video search and mining is a rapidly evolving discipline whose aim is to capture interesting patterns in video data. It has become one of the core areas in the data mining research community. In comparison to other types of data mining (e. g. text), video mining is still in its infancy. Many challenging research problems are facing video mining researchers.

Motion Trajectory Analysis. - Object Trajectory Analysis in Video Indexing and Retrieval Applications. - Trajectory Clustering for Scene Context Learning and Outlier Detection. - Motion Trajectory-Based Video Retrieval, Classification, and Summarization. - High-Dimensional Video Representation. - Three Dimensional Information Extraction and Applications to Video Analysis. - Statistical Analysis on Manifolds and Its Applications to Video Analysis. - Semantic Video Analysis. - Semantic Video Content Analysis. - Video Genre Inference Based on Camera Capturing Models. - Visual Concept Learning from Weakly Labeled Web Videos. - Personalized Video. - Face Recognition and Retrieval in Video. - A Human-Centered Computing Framework to Enable Personalized News Video Recommendation. - Video Mining. - A Holistic, In-Compression Approach to Mining Independent Motion Segments for Massive Surveillance Video Collections. - Video Repeat Recognition and Mining by Visual Features. - Mining TV Broadcasts 24/7 for Recurring Video Sequences. - YouTube Scale, Large Vocabulary Video Annotation.