Beyond Simple Cells - NIPS07 Workshop

Beyond Simple Cells:
Probabilistic models for visual cortical processing

Workshop held at NIPS 2007, Whistler, December 7

Organizers: Richard Turner, Pietro Berkes, and Maneesh Sahani

Workshop description:

The goal of the workshop is to assess the success of computational models for cortical processing based on probabilistic models and to suggest new directions for future research. Models for simple cells are now well established; how can the field progress beyond them?

We will review important questions in the field through both theoretical and experimental lenses. We have an invited panel of experimentalists to provide the latter with the hope of inspiring new research directions of greater general neuroscientific interest.

More precisely, the issues to be discussed include:

Moving beyond a two-layer hierarchy: how can we bridge the gap between objects and Gabors?
What experimental results should we attempt to model? Are current comparisons with physiology at all relevant?
What experimental results would we like to have?
What aspects of visual input are relevant for modeling (eye movements, head movements, color etc.)? How relevant is time?
Is the cortical representation best described by energy models or generative models?
What inference and learning algorithms should we use (beyond MAP: MCMC, CD, VB, score matching)?

Modelers will present their current work in short talks. The experimentalists, on the other hand, will form a panel of experts which will drive the discussion, bringing their perspective into play, and constructively expose discrepancies with biology.

Through this process, the aim is to attack the main list of questions above, and determine which aspects of the vision problem would especially benefit from a stronger collaboration between modelers and experimentalists. In addition we will have a poster session in which attendees will be able to present state of the art work.