Transcript Temple University – CIS Dept. CIS661 – Principles of Data

CIS750 – Seminar in Advanced Topics
in Computer Science
Advanced topics in databases –
Multimedia Databases
V. Megalooikonomou
Introduction
General Overview





Multimedia data
Multimedia data applications
Access of different media types
Operations on multimedia data
Infrastructure to support operations



Content: Extracting media content; indexing;
queries
Physical storage: storing media data
Creating presentations: Creating media
presentations (in response to queries); delivering
them to users at remote sites
Media types






Text/Document
Image
Video
Audio
“Traditional” data (e.g., relations, flat files,
object bases, etc)
Video and audio differ from the other media
types (temporal info):


retrievals must be continuous
support fast-forward, rewind, pause
Multimedia DBMS


A framework that manages different types of data
potentially represented in a wide diversity of
formats on a wide array of media sources
Abilities of Multimedia DBMSs (MMDBMSs)




Uniformly query data represented in different formats
Simultaneously query different media sources and apply
traditional (classical) database operations on them
Retrieve media objects from local storage devices in a
smooth jitter free (i.e., continuous) manner
Develop a presentation of the answer generated by a
query and deliver the presentation in way that satisfies
Quality of Service
Sample Multimedia Scenario


Scenario: A police investigation of a large-scale
drug operation
Types of data:






Video: from surveillance cameras that record activities
at various locations
Audio: from legally authorized telephone wiretaps
Images: from still photographs taken by investigators
Documents: seized by the police
Structured relational data: containing background
information, bank records, etc.
Geographical information systems data: containing
geographic data relevant to the investigation
Example Images Queries for
Multimedia scenario

Query 1:


Retrieve all images from the image library
in which the person appearing in the
(currently displayed) photograph appears
Query 2:

Retrieve all images from the image library
in which Dennis Dopeman appears
Example Image Queries: Issues







Image based vs keyword-based queries
Output: a ranked list of images that are
“similar” to the query image
Similarity?
Ranking?
Efficiency in supporting these queries
Association of different attributes with
images(or parts of images)
Indexing and retrieving images effectively
Example Audio Query for
Multimedia Scenario

Query 1:


Listening to a tape containing a conversation
between individual A (person under surveillance)
and individual B (somebody meeting person A)….
Query: Find the identity of individual B, given that
individual A is Denis Dopeman
Query 2:

Find all audio tapes in which Denis Dopeman was
a participant
Example Text/Video Query


Text Query: Find all documents (from the
corpora of text documents, e.g., newspaper
archives, police dept. files on old, unsolved
cases, witness statements, etc) that deal with
the Cali drug cartel’s financial transactions
with ABC Corp.
Video Query: Find all video segments in
which the victim of the assault appears.
Heterogeneous Query



Complex queries will “mix and match” data from
different media sources
Textual example: Find all individuals who have been
convicted of attempted murder in North America and
who have recently has electronic fund transfers made
into their bank accounts from ABC Corp.
Answering this query is problematic:


May requiring accessing a wide variety of databases
belonging to different police jurisdictions, etc
ABC Corp. may have accounts in hundreds of banks
worldwide each using different formats and database
systems
Heterogeneous Multimedia Query


Find all individuals who have been photographed
with Jose Orojuelo and who have been convicted
of attempted murder in North America and who
have recently has electronic fund transfers made
into their bank accounts from ABC Corp.
This query requires:



Accessing a database containing names and pictures of
various individuals
Accessing photograph database of still images,
surveillance video database
Accessing image processing algorithms to determine
who occurs in which video/still photograph
Multimedia Research Issues:
Queries





A single language within which multimedia data of
different types can be accessed
Be able to specify combination operations across different
media types (in addition to across different relations
through “join”, “union”, “difference”, etc)
Be able to access “metadata” (describing the content of
different media sources) and “raw” data supported by
different media sources
Be able to merge, manipulate, and “join” results from
different media sources
Optimize a single query or a set of queries
Multimedia Research Issues:
Content







What is the content of a media source? Under what
conditions can content be described textually and under
what conditions must it be described directly through the
original media type?
How should we extract the content of: an image, videoclip, audio-clip, free/structured document?
How to index the results of the extracted content?
How to efficiently retrieve media data on the basis of
similarity?
How one had to design a multimedia database
How should queries be “relaxed” so that not only the
original but also “similar” queries get processed?
What are efficient algorithms for processing these queries?
Multimedia Research Issues:
Storage



How do the following (standard) storage devices work?
 Disk systems
 CD-ROM systems
 Tape systems and tape libraries
How is data laid out on such devices?
How do we design disk/CD-ROM/tape servers so as to
optimally satisfy different clients concurrently, when the
operations are:
 Playback
 Rewind
 Fast forward
 Pause
Multimedia Research Issues:
Presentations and Delivery




How to specify the content of multimedia presentations?
How to specify the form (temporal/spatial layout) of this
content?
How to create a presentation schedule that satisfies
these temporal/spatial presentation requirements?
How to deliver a multimedia presentation to users
when:




There is a need to interact with remote servers to assemble the
presentation
There are bounds on the bandwidth, load and other resources
There is a mismatch between host servers’ capabilities and
customer’s machine capabilities
How to achieve Quality of Service?