Nama
: Ketlin Azura
Class
: 4c English
Tittle
: Media Semantics
Article
: Werner Haas, Werner Bailer and Michael Hausenblas
In : JOANNEUM RESEARCH
Institute of Information
Systems & Information Management
Steyrergasse
17, 8010 Graz, Austria.
Abstract
Our institute’s scope is put on
the integration of content-based and semantic technologies into
multimedia applications. We
therefore highlight the current state-of-the-art in dealing with the Semantic
Gap and present our approach based on the experience gained from projects
focusing on real-world media data.
Introduction
The key research in media
semantics still is focused on how to bridge the Semantic Gap. Following [1],
“The semantic gap is the lack of coincidence between the information that one
can extract from the visual data and the interpretation that the same data have
for a user in a given situation”. We subscribe to this view, and discuss in the
following section different approaches how to actually deal with this situation
in applications for multimedia content retrieval and production.
So far our standard methodologies to describe and
search specific content, for example an image, video or piece of music, are
mostly utilizing “piggy-back” text technologies based workingon metadata. Model-based. This class of approaches emerged from the application
of computer
vision and image understanding research to multimedia indexing and retrieval.
On the way to Media Semantics
Projects.
Our institute focuses its
research on the integration of content-based and semantic technologies into
applications and workflows for the production, and distribution of digital
content,as well as media understanding. We head for environments in the realm
of professional movie and film productions, and likewise market research. To
pursue these goals, the institute participates in national and European
projects such as SALERO, Semedia, MediaCampaign,UAd, NM2, IP-RACINE, and K-Space.
Whereas K-Space is an example from the foundational research, with NM2 and
IP-RACINE we have two media production projects that elaborate on the question:
“How to use semantic technologies together with standard media metadata to
produce content”. The three latter mentioned projects are discussed in
detail below.
Applying the Hybrid Approach. As mentioned above, we focus on real-world multimedia applications that
typically deal with large media collections. Due to the requirements of these applications,such
as scalability, multi-modality, and heterogeneity in terms of context, we
follow the hybrid approach. Features automatically extracted by content
analysis tools are represented using MPEG-7, while domain semantics are
formalized in terms of OWL. Different methodologies of integrating these two
representations (e.g. formal-driven, feature-oriented [6]) are utilized depending
on the project’s needs.
Conclusion
In this paper we have described a
number of approaches how to overcome the Semantic Gap.
We have shown how solutions could
look like in a practical setup, and have pointed out main
activities and further
directions. Based on the experience gained from our projects, we are convinced that
the integration of multimedia content analysis and Semantic Web technologies
isnecessary in order to build next generation multimedia applications.The
currently ongoing
standardization activities will
provide the basis for these technologies in the near future.
References
[1] Arnold W. M. Smeulders, Marcel Worring, Simone
Santini, Amarnath Gupta, Ramesh
Jain. Content-Based Image Retrieval at the End of
the Early Years. IEEE Transactions on
Pattern Analysis and Machine Intelligence,
22(12):1349–1380, 2000.
[2] ISO/IEC 15938, Multimedia Content Description
Interface, 2001.
[3] Jacco van Ossenbruggen, Frank Nack, and Lynda
Hardman. That Obscure Object of
Desire: Multimedia Metadata on the Web, Part 1.
IEEE MultiMedia, 11(4):38–48, 2004.
[4] Jacco van Ossenbruggen, Frank Nack, and Lynda
Hardman. That Obscure Object of
Desire: Multimedia Metadata on the Web, Part 2.
IEEE MultiMedia, 12(1):54–63, 2005.
[5] Ben Adida and Michael Hausenblas. RDFa Use
Cases: Scenarios for Embedding RDF in
HTML. Editor’s draft, W3C Semantic Web Deployment
Working Group, 2007.
[6] Peter Schallauer, Werner Bailer and Georg
Thallinger. A Description Infrastructure for
Audiovisual Media Processing Systems Based on
MPEG-7. Journal of Universal Knowledge
Management, 1(1): 26-35, 2006.
