Y. Rui, T.S. Huang, and S. Mehrotra, Constructing Table-of Contents for videos, ACM Journal of Multimedia Systems, vol. 7, no. 5, pp. 359-368, 1999.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....the shot. An average distance between the shots was computed in the L u v color space. This determined whether or not two shots were part of one logical story unit. If two shots were found similar, all the shots in between them were also merged to construct one logical story unit. Rui et al. [6] proposed the construction of table of content for videos. A time adaptive grouping of shots was done by finding the visual similarities between them. This similarity was a function of color and activity features of the shots, weighted by their temporal locality. Shots were merged together to form ....

....that the shots constitute one scene. Therefore, the weight W (i, j) is proportional to the ShotSim(i, j) and temporal proximity of the shots. This is formulated as: W (i, j) w(i, j) ShotSim(i, j) 6) where w(i, j) is a decreasing function of the temporal distance between the shots. [6] used a linearly decreasing function of the temporal distance to compute the temporal attraction between frames. However, we found out that an exponentially decreasing function performs better. Thus the weight w(i, j) decays with the distance between the middle frames of the shots under ....

Y. Rui, T. S. Huang, and S. Mehrotra. Constructing table-of-content for videos. Multimedia Systems, 7(5):359--368, 1999.

....auditory, and textual information may be employed, here we focus on the visual content of videos for indexing and structuring videos in general domains. Examples of existing LSU detectors are based on time constrained clustering [17] the overlapping links method [6] or time adaptive grouping [11]. All approaches derive a semantic video structure from a given syntactic video structure. Automatic methods result in LSU segmentations of reasonable quality [15] even based on a single feature. A reduction in manual segmentation of as much as 69 was measured for a ten hours movie collection. ....

....was the combination of automatic results from the four types of automatic LSU segmentation methods described in [15] As an estimate for the boundary likelihood the fraction of methods that found a boundary was taken. For the initial segmentation the Hue Saturation histogram feature [1] [11] was used. In the interactive segmentation process five simple features were used, viz. intensity histogram, RGB histogram, and the amount of red, blue and black in a shot. To allow for comparison of the five di#erent features, we use the similarity score distribution [16] As thresholds we used ....

Y. Rui, T.S. Huang, and S. Mehrotra. Constructing table-of-content for videos. Multimedia Systems, Special section on Video Libraries, 7(5):359--368, 1999.

....variations in network access and user terminal properties. We use a clustering based approach to identify first the appropriate clusters, and then the key frames, for every shot. While other methods that utilize clustering methods to generate visual summaries have previously been proposed [14] [17], 18] one of the contributions of our paper is the use of cluster validity analysis for automatic computation of the optimal number of representative key frames for each shot. This key frame allocation approach offers a quantitative, unsupervised alternative to the heuristic, user driven methods ....

Y. Rui, T. S. Huang, and S. Mehrotra, "Constructing table-of-contents for videos," Multimedia Systems Journal, vol. 7, no. 5, 1999.

....abstractions. Automatic indexing of video data is desirable since manual indexing is hard and indexes that are created may di#er with respect to the indexers. Low level features, which can be extracted from video data without user intervention, have been used in automatic indexing of video data [2, 10]. However, low level features are not su#cient enough to index video data based on higher level abstractions. A spatio temporal model is proposed in [5] to model semantic information of video. Modeling events by using spatio temporal attributes of objects is performed but this can only be used ....

Y. Rui, T.S. Huang, and S. Mehrota. Constructing table-of-content for videos. ACM-Springer Multimedia Systems Journal, 7(5):359--368, 1999.

....constitutes an event in these de nitions. For instance, when several events happen in the same locale, should each event be considered a scene or should they all belong to the same scene Existing scene segmentation techniques can be divided into the techniques using only visual features (e.g. [20, 5, 22, 21]) and those using both visual and audio features (e.g. 23 25] In both categories, visual similarities of entire shots or key frames (i.e. global features such as global color histograms or color moments) are typically used for clustering shots into scenes. That is, global features of ....

....accuracy and is faster than the two techniques. Editing that alternates shots of two or more lines of actions occurring in di erent places, usually simultaneously [18] The remainder of this paper is organized as follows. In Section 2, we summarize a recent technique by Rui et al. [20]. For ease of presentation, we refer to this technique as Table of Content (ToC) in this paper. ShotWeave is presented in detail in Section 3. We report experimental environments and performance results of the two techniques in Section 4 and o ers our concluding remarks in Section 5. Due to ....

[Article contains additional citation context not shown here]

Rui, Y., Huang, T.S., Mehrotra, S.: Constructing table-of-content for videos. Multimedia Systems 7 (1999) 359-368

....Video Groups Video Shots Key Frames Time Group 6 Group 3 Group 4 I Shot Figure 3: Video Tree Structure Figure 2: Hierarchy of Video Components 2.1 Video Structuring Module In this module, a video is structured according to its contents. We divide a video into five levels of components [3]. They are video key frames, video shots, video groups, video scenes, and the whole video. The hierarchy of these components in a video is shown in Figure 2. The video components are further organized into a four level tree structure as shown in Figure 3. The video tree structure can well ....

Y. Rui, T.S. Hunag, and S. Mehrotra. Constructing Table-of-Content for Videos. In 4CMMultimedia Systems Journal, Special Issue Multimedia Systems on Video Libraries, volume 7, no. 5, pages 359-368, Sept 1999.

.... tools for the classication of video sequences into various types (sports, news, movies, commercials, FLE95,NKFH98,VL98] for the retrieval of examples similar to a given video query [FB99,JVX99] or for eOEcient video browsing using high level structuring such as macrosegmentation [RHM99,YYL96]. Once the video has been segmented into shots [BGG99,ZWZS97] the issue is to deliver an interpretation and a representation of the shot content. 1 Most approaches rely on the selection of a set of key frames and on their description using texture or color features [FT98] However, motion ....

Rui (Y.), Huang (T.) et Mehrota (S.). Constructing table-of-content for videos. Multimedia Systems, vol. 5, n 7, September 1999, pp. 359368.

....boundaries in raw video sequences. 1. Introduction Successful video segmentation is necessary for most multimedia applications. In order to analyse a video sequence, it is necessary to break it down into meaningful units that are of smaller length and have some semantic coherence. Rui et al. [6] describe a video shot as: An broken sequence of frames recorded from a single camera. It is the building block of a video. It is a physical entity and is delimited by shot boundaries. Video scenes are further defined as a collection of semantically related and temporally adjacent shots, ....

Y. Rui, T.S. Huang, and S. Mehrotra, "Constructing Table-of Contents for videos", ACM Journal of Multimedia Systems, vol. 7, no. 5, pp. 359-368, 1999.

....starts and terminates are very different, and (2) The duration of dissolve transition is typically more than 0.3 second. Keywords Video Segmentation, MPEG, Shot change detection, Dissolve, Fades, Macro block type distribution. 1. INTRODUCTION Shot segmentation [4,5,6,7,9,10] and clustering [2,8] technologies have been developed for non linear video browsing and searching during the last decade. Shot segmentation represents a process of extracting temporal information, such as frame numbers, of each shot of a video based on the transition detection, and shot clustering represents a ....

....running time of test material is about 115min 54sec long. Table.1 summarizes the performance of the presented dissolve detection algorithm. The overall performance of the algorithm is very promising for video segmentation and it can be used as a basic input for shot clustering algorithms such as [2,8]. Since most of dissolve transitions satisfies proposed algorithm ( R1] to [R3] the recall ratio is very high and the miss ratio is very low. Some of miss Figure 4. Typical macro block type distribution in B frames when a wipe transition occurs Intra Interpolated Backward Prediction Forward ....

Yong, R., T. S. Huang, and S. Mehrotra, "Constructing Table-of-Content for Videos," Multimedia Systems, 7(5), 359-368 (1999)

.... edit data about the image video we get from the content provider (e.g. stock video companies) The descriptions that are used by current state of the art image video search engines can be viewed primarily as falling into the class of visual feature description [6] 8] and semantic description [9][17][18] 22] The visual feature based approach has been to obtain and utilize discriminants (features) that are useful in conducting similarity queries for visual information. Recent efforts have focused on a few specific visual dimensions such as color, texture, shape, motion and spatial ....

....will have access to a database of descriptions of the images and videos that contains not only feature descriptions, but also semantic and media based descriptions. Another example is the use of the video event hierarchy to build a video browsing interface based on table of contents for videos [17]. We show an example XML file for this interface. event hierarchy type= LOGICAL event node id = 100 event ref= 1 In Front of the house event node id = 101 event ref= 10 Shots with actor A talking event node event node id = 105 event ref= 20 Shots with ....

Y. Rui, T. S. Huang, and S. Mehrotra, "Constructing Table of Contents for Videos", ACM J. of Multimedia Systems, 1998.

.... edit data about the image video we get from the content provider (e.g. stock video companies) The descriptions that are used by current state of the art image video search engines can be viewed primarily as falling into the class of visual feature description [6] 8] and semantic description [9][17][18] 22] The visual feature based approach has been to obtain and utilize discriminants (features) that are useful in conducting similarity queries for visual information. Recent efforts have focused on a few specific visual dimensions such as color, texture, shape, motion and spatial ....

....will have access to a database of descriptions of the images and videos that contains not only feature descriptions, but also semantic and media based descriptions. Another example is the use of the video event hierarchy to build a video browsing interface based on table of contents for videos [17]. We show an example XML file for this interface. event hierarchy type= LOGICAL event node id = 100 event ref= 1 In Front of the house event node id = 101 event ref= 10 Shots with actor A talking event node event node id = 105 event ref= 20 Shots with ....

Y. Rui, T. S. Huang, and S. Mehrotra, "Constructing Table of Contents for Videos", ACM J. of Multimedia Systems, 1998.

No context found.

Y. Rui, T.S. Huang, and S. Mehrotra, Constructing Table-of Contents for videos, ACM Journal of Multimedia Systems, vol. 7, no. 5, pp. 359-368, 1999.

No context found.

Rui, Y., Huang, T. and Mehrotra, S. Constructing table-of-content for video. ACM MSJ. 7(5). pp. 359368. 1999.

No context found.

Y. Rui, T. S. Huang, and S. Mehrotra. Constructing table-of-content for videos. ACM Multimedia Systems Journal, 7(5):359--368, 1999.

No context found.

Y. Rui, T. S. Huang, and S. Mehrotra, "Constructing table-of-content for videos," ACM Multimedia Syst., vol. 7, pp. 359--368, 1999.

No context found.

Y. Rui, T. Huang, S. Mehrotra, "Constructing table-of-content for video", ACM Multimedia Systems Journal, Special Issue Mult. Systems on Video Libraries, 7(5):359-368, Sept, 1999.

No context found.

Y. Rui, T. S. Huang, and S. Mehrotra, "Constructing table-of-content for videos," Multimedia Syst. 7, 359--368 #1999#.

No context found.

Y. Rui, T. S. Huang, and S. Mehrotra, "Constructing table-of-content for videos," Multimedia Syst., Special Section on Video Libraries, vol. 7, no. 5, pp. 359--368, 1999.

No context found.

Y. Rui, et al., "Constructing table-of-content for videos," Multimedia Systems, vol. 7, no. 5, pp. 359-- 368, 1999.

No context found.

Y. Rui, T.S. Huang, and S. Mehrotra, "Constructing table-of-content for videos," Multimedia Systems, Vol. 7, pp. 359--368, 1999.

No context found.

Y. Rui, T. S. Hunag, and S. Mehrotra. "Constructing Table-of-Content for Videos", ACM Multimedia Systems Journal, Special Issue Multimedia Systems on Video Libraries, pp.359-368, Vol. 7, No. 5, Sep. 1999.

No context found.

Rui, Y., Huang, T., Mehrotra, S.: Constructing table-of-content for videos. Multimedia Systems, Special section on Video Libraries 7 (1999) 359-368

No context found.

Yong Rui, Thomas S. Huang and Sharad Mehrotra. Constructing table-of-content for videos. ACM Journal of Multimedia Systems, 1998.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC