|
|
 |
PROJECTS
The Centre for Digital Video Processing has been and is currently
involved in many projects, some of which are outlined below.
| Current Projects |
|
CLARITY: The Centre for Sensor Web Technologies
(2008 - 2013) CLARITY is a research centre that focuses on the intersection between two important research areas-Adaptive Sensing
and Information Discovery-to develop innovative new technologies of critical importance to Ireland's future industry
base and contribute to improving the quality of life of people in areas such as personal health, digital media and
management of our environment. The overarching theme of CLARITY's research programme - bringing information to life -
refers to the harvesting and harnessing of large volumes of sensed information, from both the physical world in which we
live, and the digital world of modern communications & computing. The technology developed by CLARITY will help to empower
the citizen by taming the information overload problem currently facing individuals, helping to ensure that everyone has
access to the right information at the right time.
See www.clarity-centre.com for more information.
Funding source: Science Foundation Ireland
Content Management for User-Generated Videos
(January 2008 - December 2009) In this two-year collaborative
postdoctoral research project between BT Group plc and DCU, we develop
a set of effective and efficient audio-visual analysis tools for
indexing home movie content, with particular emphasis on investigating
some principled interaction and fusion mechanism of audio and visual
cues for robust performance. These tools will take as their starting
point existing state-of-the-art work on analysis of rushes content.
Leveraging these tools, we will develop and demonstrate an application
for indexing and managing home movie content. The focus of the
application will be on collaboration annotation and semi-automatic
editing of home movies. The application will provide groups of non-
specialist users with the ability to quickly and easily create
engaging and themed content from a large repository of unstructured
home movies.
Funding source: BT Group plc and the
Irish Research Council for Science, Engineering and Technology
(IRCSET), funded by the National Development Plan
iCLIPS: Integrating Context through Annotation and Linking in Information Retrieval for
Personal Information Archives
(October 2006 - September 2009) Developments in digital data capture and storage mean
that it will soon be possible to archive one's entire life
experiences in personal information archives, or human digital
memories (HDMs). Current search engines for the World Wide Web
output document lists ranked by combining significant measures
based on the Web's link structure and a match between document
contents and search requests. The absence of links and lower
quality of the HDM document contents means that enhanced
information retrieval methods are required. This project is
concerned with the automated annotation and linking of documents
within HDM archives to facilitate more effective searching.
See more information.
Funding source: SFI Research Frontiers Programme 2006
TRIPOD: Automatically Captioning Photographs
(January 2007 - December 2009) The primary objective of Tripod
is to revolutionise access to the enormous body of visual media. Applying an innovative multidisciplinary
approach Tripod will utilise largely untapped but vast, accurate and regularly updated sources of semantic
information to create ground breaking intuitive search services, enabling users to effortlessly and
accurately gain access to the image they seek from this ever expanding resource.
See www.ProjectTripod.org for more information.
Funding source: EU FP6 STREP
MultiMatch
(May 2006 - October 2008) On the Web, cultural heritage content is everywhere, in traditional environments such as
libraries, museums, galleries and audiovisual archives, but also in popular magazines and
newspapers, in multiple languages and multiple media. The aim of the MultiMatch project is to
enable users to explore and interact with online accessible cultural heritage content, across
media types and languages boundaries.
MultiMatch plans to develop a multilingual search engine specifically designed for access,
organisation and personalised presentation of cultural heritage information.
See www.multimatch.org for more information.
Funding source: EU FR6 STREP
K-Space: Knowledge space of semantic inference for automatic annotation and retrieval of multimedia content
(January 2006 - December 2008) K-Space integrates leading European research teams to create a Network of
Excellence in semantic inference for semi-automatic
annotation and retrieval of multimedia content. The aim is
to narrow the gap between content descriptors that can be
computed automatically by current machines and algorithms,
and the richness and subjectivity of semantics in high-level
human interpretations of audiovisual media: The Semantic Gap.
See www.k-space.eu for more information.
aceMedia: Integrating knowledge, semantics and content for user-centred intelligent media services
(January 2004 - December 2008) Future market viability of
multimedia services requires significant improvements to the tools, functionality,
and systems to support target users. aceMedia will research and develop a system,
integrating knowledge discovery and embedded self-adaptability, to enable multimedia
content to be self organising, self annotating, self associating; more readily searched
(faster, more relevant results); and adaptable to user preferences and environments
(self reformatting). See www.acemedia.org for more information. Funding source: EU SIXTH FRAMEWORK PROGRAMME, Information Society Technologies
Landmark Generation from SenseCam Images
(2006 -) The problem we address is selecting, from a (large) set of
SenseCam images plus other logged data, a representative or summary of landmarks, or
significant events from a daily, weekly or longer log. The problem arises, and is well-acknowledged,
because of the large number of images captured by a SenseCam. Our approach in this
proposal is based fusing together multiple sources of diverse information including
low-level image similarity, image similarity based on salience maps, GPS and location
information, and biometric readings. See more information. Funding source: Microsoft Research
|
| Completed Projects |
|
Adaptive Information Cluster
(2003-2008) SFI support for a team of researchers who
will be based partly at the Centre for Digital Video Processing at DCU,
partly at the NCSR at DCU, and partly at University College Dublin.
The team's research will focus on adaptive information technologies,
essentially the development of software that can filter and
'personalise' the increasing amounts of information we currently receive
through channels like the Internet or mobile phones. Potential
applications of this research include personalised advertising and
information services, which would allow companies or public service
agencies to target specific market segments with marketing or
information messages; location based services and so forth.
The significance and commercial potential of the team's research has
attracted the interest of Irish-based companies, including Ericsson,
IBM, Changing Worlds and Mitsubishi Electric Research Laboratories in
Cambridge Massachusetts, USA, who will collaborate closely with the team
to resolve key issues in bringing adaptive information technologies to
market.
The work of the CDVP in this project will be in information and event
detection from audio and from video, and on adaptive media retrieval,
including retrieval from video. The research will investigate software
systems that adaptively extract and assemble meaningful content from
diverse media. For example, traffic monitoring-using audio and video
sensors to monitor traffic flows, count pedestrian and motor car volumes
and predict congestion will be much more cost effective and reliable
than the current technology. Our work will also focus on software which
will be designed to adapt to user needs and personalise content.
See www.adaptiveinformation.ie for more information.
Funding source: Science Foundation Ireland
Interactive TVe
(October 2007 - September 2008) The goal of this project was to develop a state-of-the-art
content management application suitable for an interactive TV platform. A number of content-based techniques
are exploited to enhance the remote control-operated TV interaction.
Funding source: Samsung Electronics
A Personal Content Management System for Multimedia Blogging
(2006 - 2007) The goal of this project was to develop a multimedia content management system incorporating both CDVP and
Samsung SAIT technologies. The media domain for the management system is digital photos and digital videos,
captured from camera phones, and the emphasis was to support content sharing and content
blogging using a state-of-the-art web interface, as well as automatic content annotation by examining the
content and the context of any uploaded media.
Funding source: Samsung Electronics
Network of Excellence in
Content-Based Semantic Scene Analysis and Information Retrieval (SCHEMA)
(April 2002 - March 2004)
The aim of the proposed Network of Excellence is to bring together a critical mass
of industrial partners, end users, universities and research centres in the fields of
scene analysis and information retrieval as applied to digital video. The following
new research topics, activities and standards will be investigated: Access to the
information using query structures that come naturally to human beings, Retrieval
processes which understand the nature of the information in the database,
Content-based multimedia analysis, Copyright issues of multimedia, New methods
for multimedia access and delivery, Semantic web technologies, MPEG-7 and
MPEG-21 standards, User interfaces and human factors.
Funding source: European Commission Thematic Network
MediAssit (Tools for organizing, browsing and retrieving from a personal electronic picture collection)
(October 2003 - September 2006) The objective of this project is to provide a
commercially realistic, globally competitive solution to the problem of organizing
and accessing personal image content. This is with the aim of providing the technical,
intellectual property and human capital basis for further expansion of the Digital
Media segment of the indigenous Irish Software Industry. The outputs of this project
will be a collection of software tools and the associated IP, demonstrated in an
end-to-end system delivered in two stages. At the conclusion of this project, we will
have researched, developed and field-tested a new generation of consumer-oriented
personal photo navigation tools. Funding source: Enterprise Ireland Commercialisation
Fund 2003 - Technology Development Phase
Físchlár on a PDA
(June 2000 - )
development of interfaces to allow access to
Físchlár browsers on a PDA.
Funding source: Parthus Ltd.
Digital Video Streamer
(January 2000 - )
this project
provides us with a state of the art digital video streamer capable of storing up to 2 months
of 24 hour, 7 day MPEG video and streaming to up to 150 clients in
parallel. We use Oracle Video Server on a Sun Enterprise with 6 Processors,
4 GByte RAM and 800 GByte Raid-5 storage.
Funding source: Partial funding from Sun Microsystems Ireland
and HEA Large Item of Equipment
L'OEUVRE (Linked Objects & Events for User-driven Video REtrieval)
(October 2002 - September 2005) The project will develop a new set of techniques for automatic
indexing, browsing, searching and linking of Digital Video information.
Based on automatic identification and tracking of objects appearing in
video, this will enable a new type of interaction between people and
large collections of video material which will greatly improve access
and retrieval of information from such video libraries. Two key
components of this project are the use of the MPEG-4 standard for
encoding and subsequent analysis of video material, and the use of the
MPEG-7 standard for encoding the automatically derived description of
video material. Funding source: Enterprise Ireland ATRP2002 Programme
Two Ph.D. students working on CDVP-related topics.
(October 2003 - September 2006)
Funding source: IRCSET (Irish Research Council
for Science Engineering and Technology)
Four Ph.D. students working on CDVP-related topics.
(October 2002 - September 2005)
Funding source: IRCSET (Irish Research Council
for Science Engineering and Technology)
Físchlár-2
(October 2000 - September 2003) to research and develop
techniques to scale up Físchlár to record up to 4 channels simultaneously,
and to search, filter, alert and summarise video programmes.
Delivery platforms include mobile devices.
Funding source: National Software Directorate IP2000 Programme
Postdoctoral Research Fellow researching techniques for analysis of
the audio/visual aspect of digital video.
(October 2000 - September 2002)
Funding source: Marie Curie
Postdoctoral Research Fellowship (in conjunction with RINCE)
Postgraduate research students working on video indexing and browsing.
(October 2000 - September 2001)
Funding source: Partial
funding from Dublin City University Research Committee
Físchlár - PTV
collaboration between Físchlár and PTV to
allow personalised TV recommendation and recording, and personalised TV
listings of recorded and archived TV programmes.
(January 2000 - December 2000)
Funding source: National
Software Directorate
Development of an interface to a video indexing and retrieval system.
(June 1999 - September 1999)
Funding source: TELTEC
(Dublin City University) and EU DICEMAN Project
Initial research & development of shot boundary detection
leading to the initial Físchlár recording, analysis & browsing system.
(October 1997 - September 2000)
Funding source: National
Software Directorate Advanced Software Technology Integration (ASTI) Programme
|
|
|
|
