R. Jain and A. Hampapur, "Metadata in video databases," ACM SIGMOD, Vol. 23, No. 4, pp. 27--33, 1994.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....approach is to operate on metadata (on the raw data) of less volume, containing information more pertinent to query processing engines. Metadata describes media objects in the DB enabling access, manageability and shareability of large sets of structured and or unstructured data [1] [6]. Metadata for multimedia databases can be of several types, corresponding to the di erent aspect covered: o Content Metadata: that describe the content of multimedia objects. o Representation Metadata: This data describe the physical characteristics of a media object (e.g. resolution, color ....

Ramesh Jain and Arun Hampapur. Metadata in video databases. In Proceedings of ACM SIGMOD 1994.

....must have a non zero temporal duration. A single image extracted from the video can have meaning, but this meaning will never be perceived by the viewer when this meaning is not consistent over a set of images. At the first level of the semantic index hierarchy we defined purpose. According to [38], the purpose for which the video document is made is either entertainment, information, communication, or data analysis. Recall that we only consider video documents that are made within a production environment. Therefore, the purpose of data analysis is excluded. Genre examples range from ....

R. Jain and A. Hampapur. Metadata in video databases. ACM SIGMOD, 23(4):27--33, 1994.

....must have a non zero temporal duration. A single image extracted from the video can have meaning, but this meaning will never be perceived by the viewer when this meaning is not consistent over a set of images. At the rst level of the semantic index hierarchy we de ned purpose. According to [38], the purpose for which the video document is made is either entertainment, information, communication, or data analysis. Recall that we only consider video documents that are made within a production environment. Therefore, the purpose of data analysis is excluded. Genre examples range from ....

R. Jain and A. Hampapur. Metadata in video databases. ACM SIGMOD, 23(4):27-33, 1994.

....the intended semantic meaning of the author, and defines segments on four different levels within a semantic index hierarchy. The first two levels are related to the video document as a whole, and are based on the observation that an author creates a video with a certain purpose. According to [16], the purpose for which the video document is made is either entertainment, information, communication, or data analysis. We only consider video documents that are made within a production environment, so the purpose of data analysis is excluded. The first two levels define segments based on ....

R. Jain and A. Hampapur. Metadata in video databases. ACM SIGMOD, 23(4):27--33, 1994.

....area and thus lacks an extensive body of work to draw upon. Our assumption is that content information is available as meta data. Thus, we are addressing the problem of organizing and manipulating this information in supporting smart retrievals. There is some work in this direction, most notably [10, 1, 14]. 14] is a nice paper that we draw a lot of ideas from. However, the data model presented in this paper is rudimentary and does not attempt at large scale content representation. 6, 12, 15] deal with fine grain manipulation of video this is not directly applicable, but may provide certain ....

R. Jain and A. Hampapur. Metadata in Video Databases. ACM SIGMOD RECORD, 23(4):27--33, 1994.

....a transformation. They classified an object oriented schema into three domains: class hierarchy, nested object hierarchy, and complex object hierarchy. Then they used meta data classes to build various relations between objects, such as is part of and appearing in . R. Jain and A. Hampapur[17] proposed a video model (ViMod) based on studies of the applications of video and the nature of video retrieval requests. The features of their model include content dependence, temporal extent and labeling. A feature is said to be content independent if the feature is not directly available from ....

R. Jain and A. Hampapur, "Metadata in video databases," ACM SIGMOD Record, vol.23, no.4, 1994.

....class. 1 Introduction The ability for computers to rapidly and successfully retrieve information from image databases based on the objects contained in the images has a direct impact on the progress of the revolution in communication precipitated with the increasing availability of digital video [7]. The complexity in the very nature of two dimensional image data gives rise to a host of problems that alphanumeric information systems were never designed to handle [1] A central task of these multimedia information systems is the storage, retrieval, and management of images [10] In many ....

R. Jain and A. Hampapur, "Metadata in video databases," in Sigmod Record: Special Issue on Metadata for Digital Media, vol. 23, p. 27ff, ACM: SIGMOD, December 1994.

....a transformation. They classified an object oriented schema into three domains: class hierarchy, nested object hierarchy, and complex object hierarchy. Then they used meta data classes to build various relations between objects, such as is part of and appearing in . ffl R. Jain and A. Hampapur [7] proposed a video model (ViMod) based on studies of the applications of video and the nature of video retrieval requests. The features of their model 4 include content dependence, temporal extent and labeling. A feature is said to be content independent if the feature is not directly available ....

R. Jain and A. Hampapur, Metadata in video databases, SIGMOD Record, 23 (1994), pp. 27--33.

....of available hierarchies. Hence, the system must be able to suggest hierarchies that are appropriate for specific users or specific sessions. Furthermore, the system must permit browsing of the schema (especially the data attributes) and of any other metadata associated with the multimedia data [JH94, GFSC94] ffl Data analysis systems usually deal with spreadsheet type databases comprised mainly of numerical and statistical data. However, for many applications, the bulk of the data is not numerical but 1 As we explain later, such notions of similarity can be based on textual similarity of ....

R. Jain and A. Hampapur. Metadata in Video Databases. SIGMOD Record, 23(4), December 1994.

....of indexable content (text, images, speakers, spoken text, etc. with each scene. The Interactive Cinema team at Media Lab (MIT) studies a selection of scenes in view of the video authoring, mainly from manually created annotations [28] Automatic video indexing remains an open problem [29,30, 37, 38]. selection by concepts In the research by content, an expertise on the application field is often necessary. For example for image retrieval by the color histograms, it is necessary to know the desired color distribution. When using domain and media independent concepts, content based ....

R. Jain, A. Hampapur, "Metadata in Video Databases", SIGMOD Record, 23, 4, Dec 1994, 27-33

....i :ingredients . Besides helping you to be a perfect host there are more serious applications where it is necessary to answer queries with predicates involving set valued attributes efficiently. Some examples are keyword searches and queries in annotation databases containing information on images [6, 17], genetic, or molecular data [2, 11] Further universal 2 quantifiers can be transformed into set comparisons [8] which can now be supported efficiently. Most if not all sets found in set valued attributes are small, containing less than a dozen elements. Examples for applications where almost ....

R. Jain and A. Hampapur. Metadata in video databases. SIGMOD Record, 23(4):27-- 33, December 1994.

....no structuring information (cut points, camera positions) and this information is embedded. Moreover, current image processing and video computing techniques are not able to automatically make this structure explicit, especially in the case of more abstract characteristics [Tonomura, et al. 94] Jain, Hampapur 94] Raw multimedia data are often time dependent [Hardman, et al. 94] This must be modelled in a multimedia database and affects communication [Little, Venkatesh 94] because the synchronisation between different data streams must be preserved. Most current database management systems can cope ....

Ramesh Jain and Arun Hampapur. Metadata in video databases. ACM SIGMOD Record, 23:27--33, Dec 1994.

....a sequence of temporally related timed audio video audio video streams. While this model provides adequate supports for the editing and presentation of audio video content, it lacks in an easy retrieval of an audio video stream based on inherent items of interest or meta data. Jain and Hampapur [16] proposed a video model (ViMod) based on studies of the applications of video and the nature of video retrieval requests. The features of their model include content dependence, temporal extent, and labeling. A feature is said to be content independent if the feature is not directly available from ....

R. Jain and A. Hampapur. Metadata in Video Databases. ACM SIGMOD Record, 23(4):27--33, 1994.

....because of the growth of the global information network, is stressing the development of heterogeneous, autonomous, distributed and multimedia. For the former problem, metadata assumes importance as they allow to perform content based search on non traditional data such as images [GFS94] video [JHA94], speech [GSW94] and so on. For the latter, metadata are used in order to browse, navigate and retrieve information with the focus of attention on the information content, without relying on the structure and the organization of the single databases [SSK95] A lot of modern information systems ....

R. Jain, A. Hampapur. Metadata for Video Databases. ACM SIGMOD Record 23,4 dec.1994

....done manually. We borrow the idea of modeling salient objects and events by the object oriented technology from OVID model, then integrate multimedia temporal operations into this object oriented model. An architecture, called ViMod, for a video objectbase based on video features is proposed in [JH94] The design of this model is the result of studying the metadata characteristics of queries and video features. The algebraic video data model [WDG94] allows users to model nested video structures such as shots, scenes and sequences and to define the output characteristics of video segments. A ....

R. Jain and A. Hampapur. Metadata in video database. ACM SIGMOD RECORD, 23(4):27--- 33, December 1994.

.... has been called content based retrieval in literature [1, 2] Traditional database management techniques are only capable of managing data in alphanumeric form, and hence, in the past, image and video data have been managed using keywords, which are either metadata or brief descriptions of images [3]. However, this is extremely laborious and time consuming, especially for the management of video data, thus calling for research and development of techniques for automatic extraction of semantic indices for annotation to enable content based retrieval of video segments from large corpus of video ....

R. Jain and A. Hampapur, "Metadata in video databases," ACM SIGMOD Record, vol. 23, no. 4, 1994.

....query. 1 Introduction The ability of computers to rapidly and successfully retrieve information from image databases based on the objects contained in the images has a direct impact on the progress of the revolution in communication precipitated by the increasing availability of digital video [14]. The complexity in the very nature of twodimensional image data gives rise to a host of problems that alphanumeric information systems were never designed to handle [1] A central task of these multimedia information systems is the storage, retrieval, and management of images [24] In many cases, ....

R. Jain and A. Hampapur, "Metadata in video databases," in Sigmod Record: Special Issue on Metadata for Digital Media, vol. 23, p. 27ff, ACM: SIGMOD, December 1994.

....of temporal interval [1, 12, 16] For spatial relations, most of the modeling techniques are based on projecting objects onto a two or three dimensional coordinate system. Very little attempt has been made to formally express spatiotemporal interactions of objects in a single framework. Though in [13], spatial temporal meta data for video database is defined, yet no detailed approach has been provided for data modeling and information management. In this section, we present a model to capture both spatial and temporal semantics of video data. An important feature of the this model is that it ....

....for video data where information contents at the frame level are generated. Modeling more complex information contents, such as gloomy weather, that can be extracted via image vision processing techniques is a more challenging problem and may require color based content modeling technique [13]. To facilitate identification of spatial events, union and intersection of two intervals, within the domain of video sequence consisting of frames, are defined as follows: union(I 1 ; I 2 ) fxjx 2 I 1 x 2 I 2 g, intersection(I 1 ; I 2 ) fxjx 2 I 1 x 2 I 2 g. Similarly, the union and ....

R. Jain and A. Hampapur, "Metadata in Video Databases," ACM SIGMOD RECORD, Vol. 23, No. 4, December 1994, pp. 27-33.

....(possible anchorperson shot layouts) This is achieved by color classification for face detection. Shot classification is finalized by rule based merging of anchorperson and news footage shots into news units; a structural model of the entire news program is used to construct each unit. Metadata [18], e.g. frame number, the first and the last frames of each shot, number of shots included in each news unit, and edit information, is provided at the output index file to facilitate retrieval. The presented scheme aims to answer such queries as find news programs in a TV broadcast video, skip ....

R. Jain and A. Hampapur. Metadata in video databases. SIGMOD Rec., 23, No.4:27--33, Dec. 1994.

No context found.

R. Jain and A. Hampapur, "Metadata in video databases," ACM SIGMOD, Vol. 23, No. 4, pp. 27--33, 1994.

No context found.

R. Jain, A. Hampapur, Metadata in Video Databases, SIGMOD RECORD, 23(4), Dec. 1994, pp. 27-33.

No context found.

Ramesh Jain and Arun Hampapur. Metadata in video databases. SIGMOD Record, 23(4):27-33, 1994.

No context found.

Jain, R and Hampapur, A, "Metadata in Video Databases," ACM SIGMOD, 231MOD 1994.

No context found.

Jain R., Hampapur A, Metadata in Video Databases, ACM SIGMOD Record, Vol . 23, No. 4, December 1994, pp. 27-33

No context found.

R. Jain and A. Hampapur, "Metadata in Video Databases", SIGMOD Record, 23, 4, pp. 64-71, December

First 50 documents

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