Multimedia Data Model Research Articles

Construction and Management of a Geo-Historical Information System for an Interdisciplinary and Contributory Atlas: The Historical Atlas of Limousin

At the University of Limoges in the center of France, we started developing an Historical Atlas of the region of Limousin (AHL) in 2014. The Atlas is one part of a project to gather spatial-temporal information and historical sources about the history of the region. It offers an editorial space and cartographic interface where the regional and scientific community can interact, share, and disseminate their historical knowledge and data. As such, this project represents a close interdisciplinary exchange between historians, archeologists, geographers, GIS and data scientists, as well as varied data producers such as public actors (universities, local authorities, archives), private societies (archeology and tourist operators) and associations. This article tells the story of this dialogue and explains the interdisciplinary, multimedia and spatial-temporal data model and public interface that resulted from it.

Medical Multimedia Big Data Analysis Modeling Based on DBN Algorithm

With the development of medical multimedia analysis methods based on DBN, DBN models have gained the ability to surpass medical experts in the evaluation of multimedia in some clinical examinations. Firstly, based on the existing architecture of the Internet of Things, combined with the actual characteristics of the hospital, the medical multimedia data is accessed from the IoT support platform. Secondly, the medical multimedia data modeling and classification method based on DBN is studied and analyzed. Three network structure models, a deep belief network, a stacking automatic encoder and a convolutional neural network, were introduced and analyzed. The medical multimedia data classification modeling method based on DBN was proposed to further improve the accuracy of medical multimedia data classification. The experimental results show that compared with the traditional feature extraction based neural network classification method, the classification performance is better. Thirdly, the medical state assessment model is constructed based on the multivariate Gaussian distribution theory. To study how to use the multivariate Gaussian distribution theory to design an evaluation model that can evaluate the health status of users efficiently and accurately. Finally, using the MATLAB software platform, through the experiment and simulation of 40 groups of 8×8064-dimensional physiological big data of 32 volunteers, First, determine the optimal parameters of a set of health assessment models; then use the model to learn the characteristics of physiological parameters; finally, the state assessment model to obtain the health assessment results. The experimental results show that the feature learning model based on convolutional neural network theory can effectively extract the deep features of medical multimedia big data. The health state assessment model based on multivariate Gaussian distribution theory can effectively evaluate the health status of human body.

A nonparametric Bayesian learning model using accelerated variational inference and feature selection

Developing effective machine learning methods for multimedia data modeling continues to challenge computer vision scientists. The capability of providing effective learning models can have significant impact on various applications. In this work, we propose a nonparametric Bayesian approach to address simultaneously two fundamental problems, namely clustering and feature selection. The approach is based on infinite generalized Dirichlet (GD) mixture models constructed through the framework of Dirichlet process and learned using an accelerated variational algorithm that we have developed. Furthermore, we extend the proposed approach using another nonparametric Bayesian prior, namely Pitman---Yor process, to construct the infinite generalized Dirichlet mixture model. Our experiments, which were conducted through synthetic data sets, the clustering analysis of real-world data sets and a challenging application, namely automatic human action recognition, indicate that the proposed framework provides good modeling and generalization capabilities.

Quality-based I/O resource managementfor multimedia document database servers

Slow access to disk-based multimedia data is a major limiting factor in the performance of modern multimedia Web servers connected over broadband networks. The I/O bottleneck becomes even more pronounced for currently evolving systems handling multimedia data, such as audio and video. Retrieval of complex multimedia documents needs to be handled at two levels: I/O bandwidth management for multiple multimedia streams, and interstream and intrastream synchronization for multimedia objects constituting these documents. In this paper, based on the diverse characteristics of multimedia data, we propose efficient techniques for synchronous retrieval and delivery of such data from the storage system to the main memory of the server. We propose methods to quantify user perceived quality via quality-of-presentation (QoP) parameters. We combine QoP and Object Composition Petri Net (OCPN) multimedia data modeling to develop techniques for efficient synchronous retrieval of multimedia data. Since I/O bandwidth is a precious resource, the proposed techniques have low overhead, which is $O(n\log n + mn)$ , where m is the number of logical I/O channels and n is the total number of frames of multimedia data in a scheduling period. We simulate the relative performance of these techniques under diverse I/O conditions and determine the tradeoffs between the system resources, such as memory, bandwidth, and the improvement in QoP for multimedia applications.

A Framework for the Integration of Multimedia Data.

In this paper, we propose a unified framework, that we call UniMedia, for semanticsbased integration and management of multimedia data. The framework is an adaptation and extension of the federated database technology that has been developed during the past three decades for integrating disparate data sources of alphanumeric data. In particular, we propose a multimedia data model, that we call XML/M, which encompasses diverse types of multimedia data and captures semantic relationships among them. We develop the concept of container objects which cluster relevant multimedia data and customize them according to users’ preferences. Further, we augment the concept of container objects with a version management technique. To validate utility of the UniMedia framework, we have implemented a prototype multimedia news management system.

Constraint-based automation of multimedia presentation assembly

In this paper, we present a methodology for automated construction of multimedia presentations. Semantic coherency of a multimedia presentation is expressed in terms of presentation inclusion and exclusion constraints. When a user specifies a set of segments for a presentation, the multimedia database system adds segments into and/or deletes segments from the set in order to satisfy the inclusion and exclusion constraints. We discuss the consistency and the satisfiability of inclusion and exclusion constraints when exclusion is allowed. Users express a presentation query by (a) pointing and clicking to an initial set of desired multimedia segments to be included into the presentation, and (b) specifying an upper bound on the time length of the presentation. The multimedia database system then finds the set of segments satisfying the inclusion-exclusion constraints and the time bound. Using priorities for segments and inclusion constraints, we give two algorithms for automated presentation assembly and discuss their complexity. To automate the assembly of a presentation with concurrent presentation streams, we introduce presentation organization constraints that are incorporated into the multimedia data model, independent of any presentation. We define four types of presentation organization constraints that, together with an underlying database ordering, allow us to obtain a unique presentation graph for a given set of multimedia segments. We briefly summarize a prototype system that fully incorporates the algorithms for the segment selection problem.

Multimedia data modeling based on temporal logic and XYZ System

This paper proposes a new approach to modeling multimedia data. The new approach is the multimedia data model based on temporal logic and XYZ System. It supports the formal specifications in a multimedia system. Using this model, we can not only specify information units but also design and script a multimedia title in an unified framework. Based on this model, an interactive multimedia authoring environment has been developed.

Multimedia Document Systems in Perspectives: Guest Editorial Note

Emerging multimedia applications will require the creation, storage, access and management of large archives of complex multimedia documents. These applications target various fields including medicine, business, digital libraries, education and distance learning, manufacturing and production, CAD/CAE/CAI, etc. A multimedia document may consist of diverse data types including text, image, audio, and video. These documents are either archived at a central repository or are distributed over various servers connected by broadband networks. Management and communication of multimedia documents poses considerable technical challenges in the following areas: modeling of multimedia data, designing efficient document search mechanisms and intelligent user interface, providing access to distributed multimedia documents, broadband networking for document delivery, LAN/MAN/WAN architectures for multimedia communication, synchronization protocols, and coordination between distributed sites.

Semantic modeling and knowledge representation in multimedia databases

In this paper, we present the current state of the art in semantic data modeling of multimedia data. Semantic conceptualization can be performed at several levels of information granularity, leading to multilevel indexing and searching mechanisms. Various models at different levels of granularity are compared. At the finest level of granularity, multimedia data can be indexed based on image contents, such as identification of objects and faces. At a coarser level of granularity, indexing of multimedia data can be focused on events and episodes, which are higher level abstractions. In light of the above, we also examine modeling and indexing techniques of multimedia documents.

An Approach to a Content-Based Retrieval ofMultimedia Data

This paper presents a data model tailored for multimedia data representation, along with the main characteristics of a Multimedia Query Language that exploits the features of the proposed model. The model addresses data presentation, manipulation and content-based retrieval. It consists of three parts: a Multimedia Description Model, which provides a structural view of raw multimedia data, a Multimedia Presentation Model, and a Multimedia Interpretation Model which allows semantic information to be associated with multimedia data. The paper focuses on the structuring of a multimedia data model which provides support for content-based retrieval of multimedia data. The Query Language is an extension of a traditional query language which allows restrictions to be expressed on features, concepts, and the structural aspects of the objects of multimedia data and the formulation of queries with imprecise conditions. The result of a query is an approximate set of database objects which partially match such a query.

Providing definition and temporal structure for multimedia data

Currently, multimedia (MM) applications are prepared in an ad hoc fashion. Images or video are usually produced by the MM author. The MM application itself is composed with a specific combination of a hardware platform and authoring software, and it typically contains its own specific data. Today, MM data are usually associated with a particular platform or application, but future MM developers will use repositories of reusable data shared among various applications and across various platforms. The layered multimedia data model (LMDM) emphasisizes the sharing of data and components by dividing the process of MM application development into smaller, more manageable pieces. In particular, the LMDM emphasizes the conceptual separation, manipulation, and presentation of data. When combined with a collection of MM data, the LMDM enables description of a wide variety of MM compositions. This paper describes the first two layers of the LMDM, those concerned with data definition and with the specification of temporal structures that can be reused by various MM presentations. Several examples demonstrate the advantages of the layered paradigm: modular, reusable components and isolation of platform dependencies.

Multimedia information systems

Presents a gentle introduction to multimedia information systems. The author explores the nature of multimedia data model and information system architecture, and reviews the evolution of multimedia information systems. He also discusses data model design, query processing, and browsing support, and takes a look at some state-of-the-art prototype systems. >