|
TUTORIALS
TUTORIAL 1: DEVELOPMENT OF VOICE PORTALS
12/Oct 8:30 - 12:30
Arturo Cervera, Eduardo Carrillo, Jose Javier Samper
Universidad de Valencia, Spain, University UNAB Colombia
OVERVIEW:
A voice portal uses voice synthesis, voice recognition, and pre-recorded messages
to guide the user through the content of the portal. It permits surfing on the
web using voice commands and the service can be accessed speaking into ordinary
(or mobile) phones. In addition, DTMF (Dual Tone Multi-Frequency) tones can
be sent from the user's phone and will be detected by the server. The user that
connects to the Internet through a telephone dials a number and with the telephone
network, s/he connects to a gateway. Such gateway accepts the call (for that
purpose, there is a telephone interface connected to an E1 connection, which
is a high capacity line) and makes a call with the HTTP protocol to a content
server that gives the source document. This tutorial explains the exchange data
between the development process of Voice Portals with VoiceXML, which is a marked
language oriented to communication through voice, and the use of different server
side scripting technologies with VoiceXML.
PRESENTERS
Arturo Cervera
Systems Engineering Universidad de Valencia.
Researcher
at Instituto de Robótica Universidad de Valencia since 2000.
Eduardo Carrillo
Assistant
Profesor Universidad Autónoma de Bucaramanga, Colombia, since 1995.
Current
Researcher, Instituto de Robótica Universidad de Valencia since 2001.
Jose Javier Samper
Assistant Professor Universidad de Valencia
TUTORIAL 2: THE CHALLENGE OF SEARCHING THE WEB
12/Oct 14:00 - 18:00
Ricardo Baeza-Yates
Center for Web Research, Dept. of Computer Science,
University of Chile
OVERVIEW:
We mention the main problems when searching the Web, the main
characteristics of the Web, and how to model
it as a large text database. We exemplify some of the problems
faced by the users and we characterize them.
Next, we describe the main techniques used nowadays (search engines,
directories, Web query languages, metasearchers) and how they
work. In particular we describe the architecture of major
search engines, statistical results on Web coverage by search
engines, how inverted files are used and how the text is processed.
Next, we outline some techniques that help solving the challenges
posed by the Web, including relevance ranking using hyperlinks,
direct access compression, approximate search, visual query languages,
use of text structure, and answer visualization. We finish with the
latest trends, such as new tools, metadata, and XML. This tutorial
is based on Chapters 6 and 13 of the book Modern Information
Retrieval by Berthier Ribeiro-Neto and myself (Addison-Wesley, 1999),
and the state of the art on the problem.
PRESENTER:
Ricardo Baeza-Yates received his Ph.D. in CS from U. of Waterloo, Canada,
in 1989. During 1993, he received the Organization of American States
award for young researchers in exact sciences. In 1994 obtained the
Chilean Engineers Institute award for his research work. In 1997,
with two Brazilian colleagues obtained the COMPAQ prize to the best
Brazilian CS research article in 1996. In 2002 was the first computer
scientist incorporated to the Chilean Academy of Sciences.
He is the current president of CLEI (Latin American CS Association)
and member of the board of governors of the IEEE-CS.
Currently he is professor and chair of the CS department at the
University of
Chile, as well as director of the Center for Web Research.
Among other publications, he is co-author of Modern Information
Retrieval (Addison-Wesley, 1999) and the 2nd edition of the Handbook
of Algorithms and Data Structures (Addison-Wesley, 1991), and co-editor
of Information Retrieval: Algorithms and Data Structures (Prentice-Hall,
1992).
His research interests include algorithms and data structures, text
retrieval,
web mining, and visualization applied to databases.
He is member of the ACM, EATCS, IEEE (senior), SCCC and SIAM.
TUTORIAL 3: ALGORITHMS FOR VIDEO SEGMENTATION
12/Oct 19:00 - 23:00
Carlos H. Morimoto and Thiago T. Santos, IME/USP
OVERVIEW:
We will introduce fundamental image processing algorithms for video segmentation.One
of the major problems with current video databases is the lack of automatic
tools for indexing and retrieval of video information. Although natural speech
recognition and closed caption can be used in some applications, these are not
always available and the video must be manually annotated to allow the search
for general descriptions of scenes. This tutorial is targeted to graduate students
and professionals interested in video databases with basic computer programming
skills. The tutorial starts with an introduction of the basic tools for image
acquisition and processing, including OpenCV, Java2D e JAI (Java Advanced Imaging).
These tools can be used to implement all the image processing algorithms for
video segmentation, indexing, retrieval and visualization covered in the rest
of the tutorial.
1 - Introduction - Description of the objectives - Overview of the tutorial
- Motivation and definition of the problem: Video Indexing, video retrieval
and visualization
2 - Tools for Image Processing
2.1 Java2D e JAI
2.2 OpenCV
3 - Principles of image processing
3.1 - Binary images
3.2
- Grey Level Images
3.2.1 - Edge detection
3.3 - Color Images
3.3.1 - Color Histograms
4 - Shot Boundary Detection Algorithms
4.1 - Difference of Images
4.2 - Color based methods
4.3 - Comparison: Recall and Precision
5 - Object Segmentation Techniques
5.1 - Template matching
5.2 - Texture based methods
6 - Visualization of Video Information
6.1 - Video Mosaics
6.3 - Video Mangas
PRESENTERS
Carlos Hitoshi Morimoto. Dr. Morimoto is an assistant professor at the Department
of Computer Science of the University of S˜ao Paulo. He received his B.S. and
M.S. degrees in Electronic Engineering from the Politecnic School of the University
of São Paulo and his Ph.D. in Computer Science from the University of Maryland
at College Park.
Thiago Teixeira Santos. Mr. Santos is a lecturer at SENAC
and a Ph.D. Candidate at the Department of Computer Science of the University
of São Paulo, where he obtained his B.S. degree.
|
