Multimedia Communications 1

1
Multimedia Communications

1

• Lecturer Prof. Xinhua Zhuang
• CECS EE Departments
• University of Missouri
• Columbia, MO 65211

2
The Multimedia Experience

1

• Daily Newspaper
• TV Program
• Video-on-demand
• Video Animation
• Virtual Meeting Room
• Distance Learning
• Virtual Library
• Living Books

3
Multimedia

1

• including or involving the use of several media
  of communication, entertainment, or expression

4
Multimedia vs. Multiple Media

1

• Multiple Media unique delivery mechanism,
  unique repository or mailbox
• electronic mail
• voice mail
• data files
• Multimedia single repository
• messages with integrated text, sound, images,
  video, data files, handwriting
• single access mechanism handling all media

5
Multimedia Communications

1

• basic currency of communications switches from
  narrowband voice telephony to seamlessly
  integrated, high quality, broad-band transmission
  of multimedia signals
• basic access method switches from wireline
  connections to combinations of wired and wireless
  (copper cable, ber, cell sites, satellite, and
  electrical power lines)
• basic mode of communications expands from
  people-to-people communications to include
  people-to-machine communications

6
Driving Forces in Multimedia Communications

1

• evolution of communications and data networks
  into modern POTS (plain old telephone service)
  and PACKET (a data unit belonging to level 3 of
  the ISO (International Standards Organization)
  reference model) networks, and further into an
  integrated structure
• increasing availability of almost unlimited
  bandwidth on demand in both the office and the
  home, and eventually on road, due to
  proliferation of high speed data modems, cable
  modems, hybrid fiber-coax systems, and fixed
  wireless access systems
• availability of ubiquitous (anywhere, anytime)
  access to network via LANs, and wireline and
  wireless networks

7
Driving Forces in Multimedia Communications

1

• ever increasing amount of memory and computing
  power brought to bear on virtually any
  communications and computing systems
• proliferation of smart terminals such as
  sophisticated screen phones, digital telephones,
  multimedia PCs natively handling a wide range of
  text, image, audio, and video signals
• digitization of virtually all devices such as
  cameras, video capture devices, handwriting
  terminals, sound capture devices
• standards for interconnections and communications
  between all devices attached to the network

8
Where are we Today?

1
9
Where are we Today?

1

• POTS good for narrowband voice trac
• PACKET good for data trac
• Services are separate for POTS and PACKET
  Networks
• Control is separate for POTS and PACKET Networks

10
Tomorrow's Network

1

• intelligence inside network
• intelligence at the desktop
• intelligence at the terminal
• ubiquitous services
• works with all types of access devices such as
  telephones, PC's

11
Technology Aspects of MM Systems

1

• compression and coding of multimedia signals
  standards
• organizing, storing, and retrieving multimedia
  signals streaming (real time transmission of
  multimedia signals), layering, QoS (Quality of
  Service) issues
• accessing multimedia signals by matching the user
  to the machine GUI (graphical user interface),
  SLI (spoken language interface), media
  conversion, agents
• searching multimedia archives and databases based
  on machine intelligence text, image, speech
• browsing multimedia archives and documents based
  on human intelligence text, image, audio, video

12
Illustrative Multimedia Systems

1

• teleconferencing systems which integrate voice,
  video, application sharing, data sharing
• FusionNet service which integrates the POTS and
  PACKET networks by exploiting Internet search and
  VCR-like features for viewing video, with POTS
  access to video content
• CYBRARY digital virtual library which aims to
  provide a digital library experience which is
  better than being there live
• Pictorial Transcripts system which provides a
  content-based sampled representation of video
  over the Internet

13
Technology Assumptions

1

• multimedia processing is a lot more than
  compression and coding
• multimedia applications need to be standard-based
• handling (delivery, display) of multimedia
  signals is crucial
• user interface is critical to usability of most
  applications
• multimedia experience is shared between people
  and machines

14
Image Coding Principles

1

• spatial redundancy
• repeated patterns
• image correlations in space
• spectral correlations
• temporal redundancy
• repeated objects in video sequence
• predictable moves of objects
• take advantage of human visual system
• perceptual masking of intensity, color, texture,
  time sequence

15
Image Coders FAX

1

• Group 3 FAX simple 1-D search
• Group 4 FAX simple 2-D search
• JBIG-1 prediction based on local neighborhood
• JBIG-2 soft pattern matching on segmented regions

16
Image Coders Continuous Tone

1

• JPEG
• block-based DCT
• psychophysically based scalar quantization
• entropy coding
• JPEG-2000
• modern architecture and standard
• downloadable software
• handles a broad range of conditions

17
JPEG Performance

1

• 81 Indistinguishable
• 10.71 Excellent
• 21.41 Very Good
• 321 Good
• 641 Fair

18
Video Coders

1

• H.261, H.262, H.263 motion compensated coder for
  interframe coding
• MPEG-1 multimedia standard with specifications
  for coding, compression and transmission of
  audio, video, and data streams in a series of
  packets
• MPEG-2 multimedia standard with capability of
  compressing, coding and transmitting high
  quality, multi-channel multimedia signals over
  broadband networks
• MPEG-4 object-based approach to multimedia with
  independent coding of objects, interactive
  compression of objects, ability to integrate
  synthetic and natural objects
• MPEG-7 searching, indexing, authentication of
  large databases

19
Video Coding

1

• video conferencing H.261 for ISDN, H.263 for
  POTS
• movie storage on CDROM MPEG-1, 1.2 Mb/s for
  video, 256 kb/s for audio
• broadcast video on DVD (digital video disk)
  MPEG-2, 2-15 Mb/s for video/audio
• low bit rate telephony over POTS network MPEG-4,
  10 kb/s for video, 5.3 kb/s for voice
• HDTV with 15-400 Mb/s for video

20
Searching of Multimedia Documents

1

• text-based indexing
• structured (field-based) in databases ? SQL
  (Structured Query Language) queries
• unstructured or natural language text ? full text
  search
• uses inverted index for each word records
  document and location
• can do partial matches
• can find word variants
• information extraction from documents
• speech indexing
• full recognition of text (error prone)
• event detection (word spotting)
• speaker identification used to align speech and
  text

21
Searching of Multimedia Documents

1

• audio indexing
• start and end of speech
• musical signatures for instruments
• singing voice
• image indexing
• classification by color and brightness
  histograms, texture, extracted shapes
• hand-drawn description of image content
• figure captions, textual description of image
  content
• video indexing

22
Browsing of Multimedia Documents

1

• text-based browsing
• table of contents, index, skimming
• image-based browsing
• find image of interest, examine associated text
• scene-based browsing
• find place in video for full search or editing
• video skimming browsing
• high speed playback
• audio skimming browsing
• high speed playback

23
Multimedia Conferencing

1
24
FusionNet Concept

1

• High quality audio/video over the network
• Internet provides vast development and
  information resources for browsing, searching
• POTS provides guaranteed QoS transmission with
  security and billing
• Client needs to gain access to both the Internet
  and POTS
• Server needs to gain access to both the Internet
  and POTS

25
The CYBRARY Project at ATT

1

• virtual presence in a library
• see documents in their original form on any
  screen connected to the Internet
• create a new standard for document image
  compression (DJVU) geared towards screen display
  rather than printing
• efficient document compression allows fast page
  flipping and browsing
• OCR allows full text search and indexing

26
Pictorial Transcripts

1

• Fully automatic generation of multimedia document
  from full motion video source material
• Enables broadcasters to go on-line with no
  additional manual effort
• Generate content for network-based hosting service

27
Summary

1

• multimedia is more than multiple media
• multimedia merges computing, communications, and
  information sciences
• multimedia depends on networking, computation,
  and memory
• multimedia depends on compression, standards,
  user interfaces,
• searching, and browsing
• multimedia involves integration, systems,
  testing, and conformance testing