Semantic concept-based near-duplicate video clip detection

Abstract

The detection of near-duplicate video clips (NDVCs) is a fundamental requirement of several multi-media applications. Consequently, a strong need exists for techniques that allow identifying NDVCs. Tech-niques for NDVC detection represent a video clip with a unique set of features. Conventional video signatures typically make use of low-level visual features (e.g., color histograms, local interest points). However, low-level visual features are sensitive to transformations of the video content (e.g., blurring and mirroring). Moti-vated by the observation that content transformations tend to preserve the semantic information conveyed by video clips, this dissertation introduces a novel technique for NDVC detection, leveraging model-free semantic concept detection. In particular, we realize model-free semantic concept detection by taking advantage of the collective knowledge in an image folksonomy (i.e., an unstructured collection of user-contributed images and tags), facilitating the use of an unrestricted and dynamic concept vocabulary. Experimental results obtained for the MIRFLICKR-25000 image set (used as a source of collective knowledge) and the TRECVID 2009 video set (used to create query and reference video clips) demonstrate that model-free semantic concept detection can be successfully used for identifying NDVCs.