Call for Papers
 
                       IEEE Workshop on 
        Multi-View Modeling & Analysis of Visual Scenes
 
                         June 26, 1999
                     Colorado State University
                    Fort Collins, Colorado, USA
            (in conjunction with the CVPR'99 conference)
 
                           Web Site:
           http://www.cs.rochester.edu/meetings/mview99
 
                          Organizers:
 
         Kyros Kutulakos, University of Rochester, USA
          Amnon Shashua, The Hebrew University, Israel
 
                       Program Committee:
 
  P. Anandan (Microsoft)             Jitendra Malik (UC-Berkeley)
  Olivier Faugeras (INRIA, MIT)      Tomaso Poggio (MIT)
  Michael Cohen (Microsoft)          Steve Seitz (CMU) 
  Chuck Dyer (UW-Madison)            Richard Szeliski (Microsoft)
  Michal Irani (Weizmann Institute)  Seth Teller (MIT)
  Ramesh Jain (UC-San Diego)         Demetri Terzopoulos (U. Toronto)
  Takeo Kanade (CMU)                 Andrew Zisserman (Oxford U.)
 
 
Scope & Rationale
-----------------
 
The purpose of this workshop is to bring together researchers from the
communities of computer vision and graphics interested in the
representation and manipulation of static or dynamic 3D scenes from
multiple views. In recent years, there has been a fast and noticeable
progress along the following paths: (i) image-based rendering
techniques have been developed in the computer vision and graphics
communities showing the promise of creating photo-realistic
simulations of real visual scenes, (ii) the area of
structure-from-motion has evolved into a body of results related to
the representation of 3D space and its camera views in terms of
collections of functions representing both the geometric and
photometric variability of image space, and (iii) advances in
multi-view stereo reconstruction have led to impressive results in
recovering geometrically- and photometrically-complex shapes from
large collections of images. Taken together, the fast progress along
those lines are leading to new ways to formulate the traditional
problems of structure-from-motion, motion estimation, reconstruction,
and image-synthesis.
 
Following the successful workshop on "visual scene representations"
(after ICCV'95, Boston), the "SMILE" workshop (after ECCV'98,
Freiburg, Germany), and the 1998 Image-Based Modeling & Rendering
workshop (at Stanford)---indicating a keen interest in the topics of
multi-view scene representation, reconstruction, and
visualization---we invite manuscript submissions on the following four
topics:
 
 
1. Large-scale stereo: Algorithms for recovering 3D scene structure
   from large collections of images (from five to several hundred)
 
2. Multi-view motion analysis: Techniques for simultaneously analyzing
   visual motion from multiple viewpoints (e.g., multi-perspective 
   real-time tracking, multi-view motion estimation). This also includes
   techniques for self-calibrating large collections of stationary or
   moving cameras
  
3. Image-based rendering and example-based synthesis: Image-based
   representations for view synthesis; scene reflectance modeling from
   multiple photographs; techniques for complex-motion mosaicing;
   techniques for transferring similarities across a collection of
   images of a class of objects onto novel images of that class;
   techniques for multi-view scene appearance manipulation 
 
4. Virtualized reality environments and applications: Systems and 
   algorithms for representing and visualizing dynamic 3D scenes; 
   augmented reality
 
Format
------
 
We expect that approximately 12 papers will be selected for
presentation at the workshop. The workshop's proceedings will be
available at the meeting.
 
Submission Information
----------------------
 
Manuscripts should be at most 8 pages long and should be submitted in
final, camera-ready format according to the CVPR format guidelines
(single spaced, two-column pages, 10-point Times-Roman or equivalent
font, 12 point interline spacing). A maximum of 8 pages will be
allocated in the proceedings for each accepted paper.  
 
Only electronic submissions will be accepted for the workshop. More
information on email/ftp submissions will be available on the workshop
web page by mid-December 1998.
 
Important Dates
---------------
 
Deadline for receipt of papers              25 January 1999
Acceptance notification by                  25 February 1999
Camera-ready papers due to IEEE-CS Press    21 April 1999
    
For Further Information
-----------------------
 
For any questions about the workshop please contact one of the
organizers:
 
 
Kyros Kutulakos (USA)
Phone  +1-716-275-6860
Fax    +1-716-461-2018
Email  kyros@cs.rochester.edu
Web    http://www.cs.rochester.edu/u/kyros
 
 
Amnon Shashua (Israel)
Phone +972-2-658-5988
Fax   +972-2-561-7723
Email shashua@cs.huji.ac.il
Web   http://www.cs.huji.ac.il/~shashua