|
|
|
|
|
|
|
Poster Abstracts (Alphabetical order of the last names)
Symposium Program | Talk Abstracts | Full Program (.pdf) | Short Program (.pdf)
|
|
October 25, 2008 (Saturday), 04:45 P.M. - 06:00 P.M.
|
|
1. Poster Presenter: Prof. Yoonsuck Choe (Texas A&M)
Internal State Predictability as an Evolutionary Precursor of Self-Awareness and Agency
Abstract
What is the
evolutionary value of self-awareness and agency in intelligent
agents? One way to make this problem tractable is to think about the
necessary conditions that lay the foundation for the emergence of
agency, and assess their evolutionary origin. We postulate that one
such requirement is the predictability of the internal state
trajectory. A distinct property of one's own actions compared to
someone else's is that one's own is highly predictable, and this
gives the sense of ``authorship''. In order to investigate if
internal state predictability has any evolutionary value, we evolved
sensorimotor control agents driven by a recurrent neural network in
a 2D pole-balancing task. The hidden layer activity of the network
was viewed as the internal state of an agent, and the predictability
of its trajectory was measured. We took agents exhibiting equal
levels of performance during evolutionary trials, and grouped them
into those with high or low internal state predictability (ISP). The
high-ISP group showed better performance than the low-ISP group in
novel tasks with substantially harder initial conditions. These
results indicate that regularity or predictability of neural
activity in internal dynamics of agents can have a positive impact
on fitness, and, in turn, can help us better understand the
evolutionary role of self-awareness and agency |
|
2. Poster Presenter: Prof. Byung Choi (Michigan Tech. Univ.)
Dynamic Membership Protocol for Epidemic Protocols
Abstract
Epidemic
protocols have two fundamental assumptions. One is the availability
of a mechanism that provides each node with a set of log(N) (fanout)
nodes to gossip with at each cycle. The other is that the network
size N is known to all member nodes. While it may be trivial to
support these assumptions in small systems, it is a challenge to
realize them in large open dynamic systems, such as peer-to-peer
(P2P) systems. This work introduces one possible solution which
addresses both problems; providing at each cycle a different set of
log(N) nodes selected randomly and uniformly from the entire network
under churn, and estimating the dynamic network size in the number
of nodes. |
|
3. Poster Presenter: Prof. Eun-Young Kang (Cal State LA)
Integrated Training Pipeline for Scientific Visualization
Abstract
Integrated
Training Pipeline for Scientific Visualization (ITPSV, or SV for
>short) is a student centered research and training program,
organized on the public science education projects in cosmology. The
poster introduces computer simulation, animation and game projects
developed and implemented via this program. |
|
4. Poster Presenter: Prof. Jihie Kim (USC)
Sentiment Analysis of a Student Q&A Board for Computer Science
Abstract
Online
discussion boards are widely used in higher education, extending the
availability of instructors, assistants, and materials to students
beyond the traditional classroom. Students’ emotions and attitudes
are discernible in messages posted to online question and answer
boards. Understanding student sentiments could help instructors
identify students with low confidence or high frustration, optimize
help-seeking, and potentially improve performance. Towards this end,
we present a set emotional speech acts that were used by students in
a university-level computer science course to express certainty and
uncertainty, frustration, tension and politeness. Using these new
labels, we coded a corpus of almost 1200 messages and analyzed the
results of annotated corpus. The work has application in
computational discourse analysis and in building student
help-seeking models for distance learning. This study complements
previous work on analyzing student discussions using rhetorical
speech acts, course topics, and problem tasks. |
|
5. Poster Presenter: Prof. Seon Ho Kim (U of the Dictrict of Columbia)
Viewable Scene Modeling for Geospatial Video Search
Abstract
Video
sensors are becoming ubiquitous and the volume of captured video
material is very large. Therefore, tools for searching video
databases are indispensable. Current techniques that extract
features purely based on the visual signals of a video are
struggling to achieve good results. By considering video related
meta-information, more relevant and precisely delimited search
results can be obtained. In this study we propose a novel approach
for querying videos based on the notion that the geographical
location of the captured scene in addition to the location of a
camera can provide valuable information and may be used as a search
criterion in many applications. This study provides an estimation
model of the viewable area of a scene for indexing and searching and
reports on a prototype implementation. Among our objectives is to
stimulate a discussion of these topics in the research community as
information fusion of different georeferenced data sources is
becoming increasingly important. Initial results illustrate the
feasibility of the proposed approach. |
|
6. Poster Presenter: Prof. Yoohwan Kim (UNLV)
HTSMA: a Hybrid Temporal-Spatial Multi-Channel Assignment Scheme in Wireless Mesh Networks
Abstract
A number of
multi-channel assignment schemes have recently been proposed to
improve the throughput of IEEE 802.11-based multi-hop wireless mesh
networks (WMNs) by utilizing multiple channels. However, excessive
system overhead and/or waste of bandwidth has prohibited achieving
high network throughput. We propose a synchronization-free, hybrid
temporal-spatial multi-channel assignment scheme in a tree-based
network topology, using a single radio for each host. The gateway
uses all the available channels in a round-robin fashion over time,
and the scheme ensures fair access to the gateway by all hosts. The
channel assignment for the non-gateway hosts is based on the
geographical location and channel availability (a spatial approach).
Adjacent regions have their own unique channels, and a channel can
be reused in remote regions without an interference. Simulation
results have shown that the network throughput is increased by as
much as 20.84%, and the packet completion latency is reduced by
44.02%. |
|
7. Poster Presenter: Mr. Seung Hoon Lee (UCLA)
Content Distribution in VANETs using Network Coding: The Effect of Disk I/O and Processing O/H
Abstract
Content
distribution in vehicular networks poses a great challenge due to
network dynamics and high speed mobility. In recent years, network
coding has been shown to efficiently support distribution of content
in such dynamic environments, thereby considerably enhancing the
performance. However, the related work in the literature has mostly
focused on theoretic or algorithmic aspects of network coding. In
this paper, we provide an in-depth analysis on the implementation
issues of network coding in wireless networks. First, we develop an
abstract model of a general network coding process and evaluate the
validity of the model via several experiments on real systems.
Second, we propose schemes that considerably improve the performance
of network coding under resource constrained environments. We
implement our overhead model in a network simulator and evaluate
these schemes in a large scale vehicular network. |
|
8. Poster Presenter: Mr. Minwoo Park (Pennsylvania State University)
Lattice Detection of Urban View of Buildings
Abstract
A novel and
robust computational framework for automatic detection of 2D lattice
in urban view is presented. 2D crystallographic group theory
provides a theoretical justification of degree-4 Markov Random Field
(MRF) for detecting lattice. The lattice units are proposed through
unsupervised clustering of interest points and voting for consistent
lattice units. The proposed lattice basis vectors and pattern
element contribute to the pair-wise compatibility and joint
compatibility functions in a degree-4 MRF. This allows us to
formulate lattice detection as an optimization problem, solved
within the MRF using Belief Propagation. Results demonstrate
significant advances over the state-of-the-art algorithm. |
|
9. Poster Presenter: Prof. Jungwoo Ryoo (PSU Altoona)
Immersive
Security Education Environment (I-SEE) Using Second Life |
|
10. Poster Presenter: Prof. Jongwook Woo (Cal State LA)
e-Business architecture with Enterprise Search Engine
Abstract
e-Business
has replaced dramatically or occupied the major part of the
traditional markets for the past 10 years. The e-Business
architecture moved from the legacy client-server architecture to
n-tier architecture since Internet and Web came out to the world.
Internet and web even create new product and revenue as companies
have invested in Information Technology seriously not as an option.
Content business is one of the examples in this new era. Content has
become intellectual property and a company or organization that has
content can make profit by providing the content to the customers
and the users. As the content increases high, there is a need for
search engine in order to provide the proper content to the user
quickly. Thus, search engine has received highlight in the content
industry, mostly portal sites. The paper introduces search engine -
especially internal search engine not web search engine - and its
fundamentals. Then, the paper proposes the e-Business architecture
with search engine. |
|
11. Poster
Presenter: Prof. Kyongil Yoon (College
of Notre Dame of Maryland)
Abstract
The
capability of extracting moving objects from a video sequence
captured using a static camera is a typical first step in visual
surveillance. This procedure is called a background subtraction
(BGS), and it uses the temporal distribution of pixel values over
the sequence of frames. Pixel based BGS can be improved by
considering the spatial coherence around each pixel, and in this
paper we present a method to enhance existing BGS methods using
spatial information from image segmentation. |
|
12. Poster Presenter: Ms. Mira Yun (George Washington University)
Dynamic Channel-Assignment and Scheduling considering Channel Switching Overhead in Wireless Mesh Networks
Abstract
Despite the
vast amount of research efforts in developing efficient channel
assignment and scheduling algorithms in multi-channel multi-radio
WMNs, none of them considers the overhead incurred from switching
radios dynamically from one channel to another into account. In a
802.11 card, the hardware switching delay is typically in the order
of a few hundreds of microseconds to a few milliseconds, and the
switching between two different bands (e.g., 5GHz for 802.11a and
2:4GHz for 802.11b/g) may result in even a larger delay. In this
poster, we take the switching delay into account in the design of
channel assignment and scheduling and present two algorithms: a
centralized one and a fully distributed one. The performance of our
proposed algorithms is analyzed using a discrete-event simulator
that we developed. The simulation results show that the network
throughout and the end-to-end delay can be significantly improved
using our algorithms. Some theoretical issues are also addressed. |
|
|
|
KOCSEA (Korean Computer Scientists & Engineering Association in America) |
