2nd IEEE
Workshop on Vision for Human Computer Interaction (V4HCI)
New York,
NY
June 22nd,
2006
This
workshop is aimed towards bringing together computer vision and pattern
recognition (CVPR) scientists whose work impacts human computer interaction
(HCI). This workshop grow from earlier
workshops on CVPR for HCI that were held in Madison (in 2003) and in Washington
DC (in 2004). In 2005, the two earlier workshops were merged together to create
V4HCI. Following last year’s success, we seek to attract contributions from a
wide range of theoretical and application areas. In doing so, we hope to facilitate discussions and debates that
will identify the opportunities and challenges in applying CVPR to HCI.
Relevant topics for the workshop include but are by no means
limited to:
· Affective computing
· Brain-computer interface
· Database for training and testing
· Event detection and recognition
· Facial expression recognition
· Human motion and gesture recognition
· Haptic interface
· Learning
· Multimodal interface and integration
· Perceptual user interface
· Sign language analysis and recognition
· User, context, and task modeling
Papers should not exceed 6 double-column pages. Only
original manuscripts will be considered.
Papers should be written within the context of HCI, and have
direct implications and contributions to a specific HCI problem. All areas of HCI will be considered subject
to the constraint that the submission makes an important contribution to
computer vision and/or pattern recognition techniques. In general, papers that
solely describe a computer vision or pattern recognition approach with
potential applications to HCI or other areas should be submitted to the CVPR
general conference. Note that although applications of known pattern
recognition and computer vision techniques are welcome, we will give priority
to those works that also make theoretical contributions to PR or CV. Papers
focusing on applications should justify the PR or CV approach chosen for the
problem under consideration. Experimental results should be consistent with
such claims. A section with a direct discussion of the results is encouraged.
Please use CVPR’s Latex style, available here.
Papers should not identify the authors.
Survey papers are welcomed and encouraged. Authors interested in
submitting a survey article may want to contact the program co-organizers
(v4hci@ece.osu.edu) prior to submission.
Double
submissions (or articles describing the same approach and similar results) to
both the main conference and the workshop are not permissible.
Workshop
Chairs
Aleix
M. Martinez
Ohio
State University, Department of Electrical and Computer Engineering
Larry
S. Davis
University
of Maryland, Department of Computer Science
Program
Committee
Jake Aggarwal, University of Texas
Chuck Anderson, Colorado State University
Alberto
del Bimbo, University of Florence
Rama
Chellappa, University of Maryland
Sven
Dickinson, University of Toronto
Tom
Huang, UIUC
Avi
Kak, Purdue University
Branislav Kisacanin,
Delphi
Mathias Kölsch, UCSD
Michael Lyons, ATR
Maja Pantic, Imperial
College London
Vladimir Pavlovic,
Rutgers
James Rehg, Georgia Tech
Stan Sclaroff, Boston University
Lynne A Slivovsky, Cal Poly State University,
Carlo Tomasi, Duke University
Ming-Hsuan
Yang, Honda Fundamental Research
Jordi
Vitria, CVC, UAB
Keynote Speakers
Sudeep Sarkar, University of South Florida.
Title:
Moving Vision for Human Computer Interaction Forward
Abstract: Three fundamental vision problems that are crucial in
the context of unaided V4HCI are (a) segmentation, grouping, and tracking of
hands, persons, and faces, (b) going beyond single gestures, and (c)
integration of faces and gestures. As examples, we will share our ideas and
thoughts on these issues using the context of unaided sign language
recognition. We will follow this with a proposal to set up a V4HCI Challenge
Problem framework. It would include (a) data sets, (b) study and experiment
definitions, and (c) baseline algorithms to benchmark minimum expected
performance. By drawing parallels from face and gait recognition research we
will see how benchmarks and challenge problems can be used to direct and focus
research towards eventual transition to operational scenarios.
Alex Pentland, MIT.
Title:
A Computational Model of Social Signaling
Abstract: I have proposed that unconscious face, hand, and body
gestures form a visual motion texture that conveys social signals, and are an
important determinant of behavior. In this talk I will survey the results
obtained using this computational model for the perception of social
displays. Results include accurate prediction of dating outcomes,
business card trading, persuasiveness of elevator pitches, requests for
additional information, ratings of interest, and salary negotiation results.
Schedule
Opening Remarks
8:20 am Welcome address
Session I: Keynote Speaker
8:30am Sudeep Sarkar
Moving
Vision for Human Computer Interaction Forward
Session II: Human Interaction
9:30am Tracking
of Multiple Humans in Meetings
B. Wu and
R. Nevatia
9:50am
wikiTable: finger driven
interaction for collaborative knowledge-building workspaces
S.
Baraldi, A. Del Bimbo, L. Landucci and A. Valli
10:15 am – 10:35 am Coffee Break
Session III: Recognition
10:35am Semi-autonomous Learning
of Ojbects
H. Kim, E.
Murphy-Chutorian, and J. Triesch
11:55am Three-dimensional Reconstruction of shape and
motion for the analysis of American Sign Language
L. Ding
and A.M. Martinez
11:15am Dynamical Motion Vocabularies
for Kinematic Tracking and Activity Recognition
O.C.
Jenkins, G. Gonzalez and M.M. Loper
11:35am Open Hand Detection in a Cluttered Single Image using Finger
Primitives
M. Baris Caglar and
N. da Vitoria Lobo
12:00 pm – 1:30
pm Lunch
Session IV: Keynote Speaker
1:30pm Alex Pentland
A
Computational Model of Social Signaling
Session V: Faces
2:30pm Fully
Automatic Facial Action Unit Detection and Temporal Analysis
M.
Valstar and M. Pantic
2:50pm Facial Action Coding
Using Multiple Visual Cues and a Hierarchy of Particle Filters
J.
McCall and M. Trivedi
3:15pm – 3:30 pm Coffee break
Session VI: Faces (cont)
3:30pm Kernel-based Recognition of Human Actions
Using Spatiotemporal Salient Points
A.
Oikonomopoulos
3:50pm Robustifying Eye Interaction
D.W.
Hansen
Session VII: Poster Session
Starts: 4:15pm
A Comprehensive Empirical Study on
Linear Subspace Methods for Facial Expression Analysis
C. Shan, S. Gong, and P.W. McOwan
Multi-view Appearance-based 3D
Hand Pose Estimation
H. Guan, J. Silk Chang, L.
Chen, R.S. Feris and M. Turk
Robust Online Change-point
Detection in Video Sequences
G. Tsechpenakis, D. Metaxas, O.
Hadjiliadis, and C. Neidle
Performance Evaluation of
Vision-Based High DOF Human Movement Tracking: A Survey And Human Computer
Interaction Perspective
D. Heckenberg
Active Learning in Face
Recognition: Using Tracking to Build a Face Model
Belief Propagation Driven Method
for Facial Gestures Recognition in Presence of Occlusions
W. Liao and I. Cohen
Rapid Signer Adaptation for
Isolated Sign Language Recognition
U. von Agris, D. Schneider, J. Zieren and K.-F. Kraiss
6:00pm Closing remarks
Important Dates
Workshop:
June 22, 2006
Submissions, questions, and requests
Please submit your paper by email
to v4hci@ece.osu.edu