MODVIS 2013 is a small workshop dedicated to the investigation of formal models in vision research.
The workshop will be held at the Edgewater Beach Hotel in Naples, Florida. This is the sister property to the main VSS hotel, the Waldorf Astoria (formerly the Naples Grande). Reservations may be made through the VSS web site at the conference group rate of $189/night. (Resort fees and parking charges will be waived for VSS attendees.)
MODVIS 2013 will be held immediately preceding the VSS annual meeting, all day on Wednesday and Thursday (May 8-9), and the morning of Friday May 10th.
Why do we need yet another conference?
We intend this to be a very special workshop, one that fills a critical niche. Specifically, we hope that it will:
- Help move our field forward, because substantial progress in any field is not possible without formal theories. Modeling work at VSS and similar conferences is scattered among the diverse parallel sessions, making it nearly impossible to keep up with new developments. Our proposed format will bring a diverse group of modelers together.
- Facilitate interactions among theoretically-minded researchers. A discussion among modelers always goes beyond phenomena and effects. Many papers at ordinary meetings rarely go beyond the data. Our intent is to focus on the role of mechanisms (vision algorithms).
- Encourage people to present their formal theories and models in some detail (presentations may be slightly longer than at VSS). We all would like to see more equations and cost functions and to be able to discuss the stability and complexity of our models. Time will be available to do this within the sessions and on the beach between them. Models used in only one part of the field could prove to be very useful in another.
- Attract machine vision people interested in human vision, again, because our current knowledge of human vision has finally reached the point where our models can actually be used by seeing machines.
- Keep us up-to-date with modeling across all specialized areas within vision. This will be beneficial for those of us who teach. Note that one could offer a graduate seminar in vision every year based on the talks that will be presented at the kind of conference we are proposing.
- Help integrate vision into a single field: We think that solving the obvious binding problem in our field should be possible, and even helpful, for solving the binding problem in the brain.
Jeffrey B. Mulligan
NASA Ames Research Center
Due to the small and intimate nature of the workshop, we have not deemed it necessary to form a program committee; but if you are interested in helping out in any way, shape or form, please let us know.
May 8 (Wednesday)
||Qasim Zaidi and Elias Cohen
Model for the salience of mirror symmetry in natural patterns.
|9:30-10:00||Ingo Frund and James Elder
Human selectivity for statistical properties of natural shapes.
|10:30-11:00||Vicky Froyen, Jacob Feldman and Manish Singh
Perceptual grouping as Bayesian estimation of mixture models.
Publication bias, modeling, and theorizing in vision science.
How do object reference frames and motion vector decomposition emerge in laminar cortical circuits?
|2:30-3:00||Bo Cao and Arash Yazdanbakhsh
A neural model of MSTd neurons for eye movement compensation.
A biophysical model of retinal circuitry: spatiotemporal dynamics.
Oliver Layton and Arash Yazdanbakhsh
May 9 (Thursday)
||Michele Rucci, Jonathan Victor and Murat Aytekin
Seeing space through time during natural fixation.
|9:30-10:00||Emre Akbas and Miguel Eckstein
Object detection with a foveated visual system.
Perceptual weighting of local contrast information in an edge integration model of neural lightness computation.
What is our baseline model of visual processing?
|2:00-2:30||Holly Gerhard, Felix Wichmann and Matthias Bethge
How Sensitive Is the Human Visual System to the Local Statistics of Natural Images?
Modeling ‘Binoculomotor’ Dynamics.
|3:30-4:00||Harald Ruda and Ennio Mingolla
A model of depth ordering from motion occlusion and disocclusion in textured scenes.
|4:00-4:30||Inna Tsirlin, Laurie Wilcox and Robert Allison
A biologically-plausible computational model of depth from monocular occlusions and binocular disparity.
May 10 (Friday)
|9:00-9:30||Jian Ding, Stanley Klein and Dennis Levi
Modeling studies on binocular combination.
|9:30-10:00||Stanley Klein, Thom Carney, Claudio Privitera, Austin Roorda, Ram Sabesan and Gene Switkes
Modeling perception from single and multiple cone stimulation.
A geodesic on a non-differentiable part of a 3D surface and its role in visual perception.
|11:00-11:30||Ilan Kadar and Ohad Ben-Shahar
From Perceptual Relations to Scene Gist Recognition.
The abstract submission site is a little quirky, so we have provided detailed instructions to help you make it through the process without mishap. We suggest that you either print these instructions, or keep this page open while you submit your abstract in another window.
Abstracts will be accepted on a first-come first-served basis; slots on the program will be guaranteed by payment of the registration fee.
1. Click the "SUBMIT ABSTRACT" button below.
2. You will be redirected to a different server, which handles abstract submission. You may receive a message saying "Session Timeout - there was a problem handling your request." If that happens, click the link which says "Return to Application."
3. You will be redirected to the sign-in page. If this is your first visit, you will need to create an account here. Otherwise, enter your email address and password, and click the button labeled "Sign In To Account".
4. You will be redirected to the "Presenter Information" page. In the section labeled "Submitter account" click the link labeled "Add myself as presenter".
5. Now a yellow block should appear in the section labeled "Submit on behalf of someone else". If you have additional co-authors, add them by clicking the link labeled "Add Presenter".
5b. If you clicked "Add Presenter", you will be taken to a page labeled "Create Account" - ignore all text related to account creation and email addresses, and just enter the name and affiliation of your co-author.
5c. If you added a co-author, change their "Type" from "Presenter" to "Co-Presenter".
6. After all co-authors have been added, click the button labeled "Continue".
7. You will be redirected to a page labeled "Presenter Information". Select the most appropriate topic for your presentation from the pull-down menu labeled "Topic".
8. Enter your title in the field labeled "Title", and abstract body in the field labeled "Topic Text". There is a 150 word limit for the abstract, but unfortunately if you are over you will not be told the number of words, only that your are over.
9. For "Presentation Type" select "Lecture".
10. Enter a list of up to 5 comma-separated keywords.
11. Click the "Continue" button.
11b. Correct any errors and click "Continue".
12. You will be redirected to a page labeled "Presentation Review". Check all of your information for correctness. When all is correct, click "Save Changes".
13. You will be redirected to a page labeled "Instruction to Submitters". Please note that while the system allows multiple submissions, each MODVIS participant may only submit a single abstract for presentation (although you may be a co-author on other presentations).
14. Congratulations! You are done, and may click the link labeled "Click here to log out ..."