[Date Prev][Date Next][Date Index]
CSLI Calendar, Wednesday, 11 December 2002, vol. 18:14
CSLI CALENDAR OF PUBLIC EVENTS
______________________________________________________________________
11 December 2002 Stanford Vol. 18, No. 14
______________________________________________________________________
A weekly publication of the
Center for the Study of Language and Information (CSLI)
Stanford University, Ventura Hall, Stanford, CA 94305-4115
http://www-csli.stanford.edu/
____________
ACTIVITIES FROM 11 DECEMBER 2002 TO 3 JANUARY 2002
WEDNESDAY, 11 DECEMBER 2002
3:45pm Psychology Department Colloquium
Jordan Hall 420:041
Culture and Point of View
Richard Nisbett
University of Michigan
http://www-psych.stanford.edu/news.html
THURSDAY, 12 DECEMBER 2002
4:00pm PARC Forum
George Pake Auditorium at PARC
Nexus Analysis: Expanding the Circumference of Discourse Analysis
Ron Scollon and Suzanne Scollon
Linguistics, Georgetown University
http://www.parc.com/forum/
Abstract below
4:00pm BISC Seminar
405 Soda Hall, Berkeley
Mining Colors with Fuzzy Decision Trees
Marcin Detyniecki
CNRS, France
http://buffy.eecs.berkeley.edu/Seminars/
Abstract below
4:15pm CSLI Seminar on Computational Learning and Adaptation
Cordura Hall, room 100
Hierarchies of Models: Toward Understanding Planetary Nebulae
Kevin Knuth
Computational Sciences Department, Code IC NASA Ames Research Center
http://cll.stanford.edu/scla.html
Abstract below
4:15pm Fundamental Themes in Neuroscience Seminar
Munzer Auditorium, Beckman B060
Synaptic Mechanisms for Cortical Map Plasticity in Rat Somatosensory
Cortex
Daniel Feldman
UC San Diego
http://www-med.stanford.edu/sbrc/calendar/
WEDNESDAY, 18 DECEMBER 2002
10:00am UC Berkeley Dissertation Defense
310 Soda Hall (UC Berkeley)
An Empirical Approach to Grouping and Segmentation
David Martin
http://buffy.eecs.berkeley.edu/Seminars/
http://www.cs.berkeley.edu/~dmartin/
Abstract below
THURSDAY, 19 DECEMBER 2002
4:15pm Fundamental Themes in Neuroscience Seminar
Munzer Auditorium, Beckman B060
New Tricks for an Old Molecule: NCAM's Role in Presynaptic Maturation
Lynn Landmesser
Case Western
http://www-med.stanford.edu/sbrc/calendar/
____________
Stanford Blood Center status: Critical shortage of O- and A- and a
shortage of O+, AB-, and A+. For an appointment:
http://bloodcenter.stanford.edu/ or call 650-723-7831. It only takes
an hour of your time.
____________
ANNOUNCEMENT
Please note that the CSLI Calendar will not appear on December 18 and
25 and on January 1. Note that CSLI itself will be closed from
December 23, 2002 to January 1, 2003 inclusive.
____________
PARC FORUM
on Thursday, 12 December 2002, 4:00pm - 5:00pm
George Pake Auditorium, PARC
http://www.parc.com/forum/
Nexus Analysis: Expanding the Circumference of Discourse Analysis
Ron Scollon, Professor of Linguistics
Suzanne Scollon, Research Associate
Georgetown University, Department of Linguistics
At any moment of action, many discourses are present, most of them
buried in practices of the social actors or as 'histories' in the
objects and places present in the action. In analyzing how we lit a
camp stove to make breakfast, introduced email to university distance
delivery 20 years ago, and traced Cold War ideological cleavages
within a group of friends who exercised together in a public park
during the Taiwan Missile Crisis, we show how a nexus analysis, as a
form of discourse analysis, can give us a fresh view of moments of
social interaction. Such a view gives us leverage in bringing about
change in the discourses that emanate from human action.
Discourse analysis has been a productive line of research for the past
four decades or so at both the micro-social interactional level and at
the level of the study of macro-socio-political forces in our lives.
Mediated Discourse Analysis (MDA) is a form of sociocultural
(activity/practice) analysis that seeks to clarify the many complex
relations between discourse and social action. Nexus analysis is a
way of opening up the circumference around moments of human action to
begin to see the lines, sometimes visible and sometimes obscured of
historical and social process by which discourses come together at
particular moments of human action as well as to make visible the ways
in which outcomes such as transformations in those discourses, social
actors, and mediational means emanate from those moments of action.
We show with three examples how we engage in a nexus of practice by
recognizing a zone of identification, how we navigate the nexus of
practice, and how we change the nexus of practice, taking into
consideration the hidden discourses, the hidden dialogicality, that
influence moments and outcomes of action.
About the Speakers: Suzanne Scollon's research in the field of
linguistics has explored socialization to language and worldview in
several multilingual societies including Fort Chipewyan, Alberta,
Honolulu and Hong Kong. She is particularly interested in the
sociocultural psychology node of her and Ron Scollon's nexus of
practice approach. Ron Scollon's research has encompassed the study of
language use and language structure, and more recently, mediated
discourse analysis, geosemiotics, discourse and ethnography, and
activity/practice theory.
Their current work includes several books: Mediated Discourse: The
Nexus of Practice (2001, Routledge) examines the relations between
discourse and a nexus of practice and ways of explicating those
linkages; Discourses in Place: Language in the Material World
(estimated publication 2003, Routledge) discusses a kind of
geosemiotics -- the integration of social interactionist theory
(including all forms of spoken discourse), visual semiotics (and
significantly including text as fixed and therefore visual forms), and
'place semiotics', especially the built environment; and Nexus
Analysis: Expanding the Circumference of Discourse Analysis (in
preparation) develops the concept of nexus analysis as an activist
theory of participation in social change through discourse analysis.
In addition to their research, Ron Scollon and Suzanne Scollon have
taught at universities across the world including the University of
Hong Kong and the University of Central Finland. They have also
consulted with a large list of organizations such as Nokia Mobile
Phones and the National Endowment for the Humanities on matters of
intercultural communication and the intercultural aspects of the
introduction of technology.
____________
BERKELEY BISC SEMINAR
on Thursday, 12 December 2002, 4:00pm-5:30pm
405 Soda Hall (Berkeley)
http://buffy.eecs.berkeley.edu/Seminars/
Mining Colors with Fuzzy Decision Trees
Marcin Detyniecki
CNRS, France
On the one hand, in the recent years, knowledge discovery from data
(and data mining) introduces new methodologies to extract and discover
automatically knowledge from data repositories. On the other hand, the
growth of multimedia data has caused a corresponding need to analyze
and exploit it. Thus, it appears natural and promising to link data
mining with multimedia data. Unfortunately, dealing with multi-media
introduces a new problem related to the polymorphism of the data
(video, texts, images, sounds, temporal data, metadata, etc). A
solution lies in providing a flexible and automated data-mining tool,
which will induce knowledge from all kinds of data. A particular
instance of such tools is the fuzzy decision tree based algorithm.
PLAN OF THE PRESENTATION The talk is dived in two parts:
- In the first part, a quick "tutorial" and the state of the art for
the Fuzzy Decision Trees will be presented.
- The second part addresses the problem of discovering knowledge from
images using the Fuzzy Decision Trees. We present how we discover
three indexing rules for video-news. These rules are intended for
finding the structure (macro-segmenting) the video news. We use a
fuzzy decision tree to learn these rules directly from simple color
proportions extracted from key-frames taken from the video.
About the speaker: Dr. Marcin Detyniecki is currently researcher of
the French National Science Foundation (CNRS - Centre national de la
recherche scientifique). He obtained BS and MS in physics and
mathematics at the University of Paris 6. He received in 2000 his
Ph.D. in Artificial Intelligence from the same university. During 2001
he held postdoctoral position at the Berkeley Initiative in Soft
Computing (BISC) at the University of California, Berkeley. His work
focuses on theoretical and practical aspects of data mining, of
aggregation and fusion of imprecise and uncertain information and of
multimedia systems.
____________
CSLI SEMINAR ON COMPUTATIONAL LEARNING AND ADAPTATION
on Thursday, 12 December 2002, 4:15pm to 5:30pm
Cordura 100
http://cll.stanford.edu/scla.html
Hierarchies of Models: Toward Understanding Planetary Nebulae
Kevin H. Knuth
NASA Ames Research Center,
Computational Sciences Department, Code IC
Stars like our sun (initial masses between 0.8 to 8 solar masses) end
their lives as swollen red giants surrounded by cool extended
atmospheres. The nuclear reactions in their cores create carbon,
nitrogen and oxygen, which are transported by convection to the outer
envelope of the stellar atmosphere. As the star finally collapses to
become a white dwarf, this envelope is expelled from the star to form
a planetary nebula (PN) rich in organic molecules. The physics,
dynamics, and chemistry of these nebulae are poorly understood and
have implications not only for our understanding of the stellar life
cycle but also for organic astrochemistry and the creation of
prebiotic molecules in interstellar space.
We are working toward generating three-dimensional models of planetary
nebulae (PNe), which include the size, orientation, shape, expansion
rate and mass distribution of the nebula. Such a reconstruction of a
PN is a challenging problem for several reasons. First, the data
consist of images (taken from a single viewpoint) obtained over time
from the Hubble Space Telescope (HST) and ground-based long-slit
spectra. Second, the fact that we have two disparate data types
requires that we utilize a method that allows these data to be used
together to obtain a solution. To address these first two challenges
we employ Bayesian model estimation using a parameterized physical
model that incorporates much prior information about the known physics
of the PN. As we have found that the forward problem of this
comprehensive model is extremely time consuming, we have introduced a
method relying on a hierarchical set of models, where each model in
the hierarchy allows us to estimate increasingly more detail. In this
talk, we describe these analysis techniques and explore the advantages
and disadvantages of employing such a set of hierarchical models.
This talk describes joint work with Arsen R. Hajian.
____________
UC BERKELEY DISSERTATION DEFENSE
on Wednesday, 18 December 2002, 10:00am
310 Soda Hall (UC Berkeley)
http://buffy.eecs.berkeley.edu/Seminars/
An Empirical Approach to Grouping and Segmentation
David Martin
UC Berkeley
http://www.cs.berkeley.edu/~dmartin/
I presents a novel dataset of 12,000 segmentations of 1,000 natural
images by 30 human subjects. The subjects marked the locations of
objects in the images, providing ground truth data for learning
grouping cues and benchmarking grouping algorithms. We feel that the
data-driven approach is critical for two reasons: (1) the data
reflects "ecological statistics" that the human visual system has
evolved to exploit, and (2) innovations in computational vision should
be evaluated quantitatively. We develop a battery of segmentation
comparison measures that we use to both validate the consistency of
the human data, and to provide approaches for evaluating grouping
algorithms. In conjunction with the segmentation dataset, the various
measures provide "micro-benchmarks" for boundary detection algorithms
and pixel affinity functions, as well a benchmark for complete
segmentation algorithms. Using these performance measures, we can
systematically improve grouping algorithms with the human ground truth
as our goal. Starting at the lowest level, we present local boundary
models based on brightness, color, and texture cues, where the cues
are individually optimized with respect to the dataset, and then
combined in a statistically optimal manner with classifiers. The
resulting detector is shown to significantly outperform the prior
state-of-the-art. Next, we learn from data how to combine the boundary
model with patch-based features in a pixel affinity model to settle
long-standing debates in computer vision with empirical results: (1)
brightness boundaries are more informative than patches, and vice
versa for color; (2) texture boundaries and patches are the two most
powerful cues; (3) proximity is not a useful cue for grouping, it is
simply a result of the process; and (4) both boundary-based and
region-based approaches provide significant independent information
for grouping.
____________
END MATERIAL
The CSLI Calendar appears weekly on Wednesdays throughout the academic
year. Announcements, abstracts, and other information to appear in
the Calendar should be submitted to the editor, who reserves the right
to decide what does or does not go in the calendar
mailto:incalendar@csli.stanford.edu
Requests to be added to the mailing list should be sent to
majordomo@csli.stanford.edu. With the lines in the body of the text
of either
subscribe csli-calendar
for the long form or
subscribe csli-short-calendar
for the short form (i.e., no abstracts). Problems with subscribing or
unsubscribing should be sent to
owner-csli-calendar@csli.stanford.edu.
The full current issue is at
http://www-csli.stanford.edu/Archive/calendar/current.shtml
and the archives at
http://www-csli.stanford.edu/Archive/calendar/
People on most of the CSLI computers can type 'help csli-calendar' to
see the current issue.
The CSLI Calendar is also posted each week to
news://nntp-csli.stanford.edu/csli.bboard.
and
news://news.stanford.edu/su.events
Information about CSLI's research program is available at
http://www-csli.stanford.edu/
For maps to the Stanford University campus see
http://www.stanford.edu/home/visitors/maps.html
____________