Citation Context

...ven clip, the identification of recurrent instances of a given clip, and similar video content search in a sense of coarse or clear semantic meanings. Application comprises video copyright management =-=[6]-=- [13], video content identification [1] [6] [9] [12] [15] and content-based video retrieval [2] [3] [4] [14]. For content-based video retrieval and identification, most existing search methods are der...

Citation Context

...rity between two video sequences is measured by identifying the closest match for every single frame in each sequence. The time complexity is linear to the product of the lengths of the two sequences =-=[6]-=-. Obviously, such a process is impractical for very large video databases. One promising way to reduce the computational cost is to construct a compact summary which is described by much fewer keyfram...

Citation Context

...erred to as signatures or fingerprints which summarize the global statistic of low-level features. Typical features include color, motion and ordinal signature [8], [38] and prototype-based signature =-=[5]-=-, [6]. These global feature based approaches are suitable for identifying almost identical videos, and can detect minor editing in the spatial and temporal domain [5], [6], [8], [9], [38]. Our analysi...

Citation Context

... To detect similar video content on the web, we require an algorithm to be capable of measuring sequences of any length, be robust against temporal re-ordering, and extremely efficient to execute. In =-=[1]-=-, we have proposed a randomized algorithm called video signature (ViSig), aimed at measuring the fraction of visually similar frames shared between two sequences. ViSig produces a compact representati...

Citation Context

...is definitely more concise and convenient than query-by-shot. Existing approaches in clip-based retrieval include [1-9, 12, 16, 17]. Some researches focus on the rapid identification of similar clips =-=[2, 3, 5, 7, 12]-=-, while the others focus on the similarity ranking of videos clips [1, 4, 6, 8, 9, 16, 17]. In [2, 3, 7], fast algorithms are proposed by deriving signatures to represent the clip contents. The signat...

Citation Context

...dancy detection. 2.2 Multimedia Information Retrieval Shot similarity and clip similarity have been extensively addressed for retrieval and clustering. Previously, clip and video copy detection (e.g. =-=[5, 11]-=-) were investigated by using image similarity measure with low-level global features, for instance, color histograms. Fast signature-based methods (e.g. [5]) were proposed to identify similar clips, w...

Citation Context

...matching algorithms have been proposed. Generally speaking, global features are suitable for identifying the majority of copies in formatting modifications such as coding and frame resolution changes =-=[7, 8, 11, 12, 18, 35]-=-, while segment or shot-level features can detect some of copies with simple to moderate level of editing [35]. More sophisticated approaches normally involve the intensive use of feature matching at ...

Citation Context

...approaches in clip-based retrieval include [1-14]. Some researches focus on the rapid identification of similar clips [1-6], while the others focus on the similarity ranking of video clips [7-14]. In =-=[1, 2, 4, 6]-=-, fast algorithms are proposed by deriving signatures to represent the clip contents. The signatures are basically the summaries or global statistics of low-level features in clips. The similarity of ...