V1/MT model, MacIntosh application

David Heeger (heeger@violet.Stanford.EDU)
Thu, 12 Oct 95 11:26:59 PDT

The V1/MT model is a MacIntosh application that serves as a tutorial
introduction to our research on computational models of visual
cortical processing. This application teaches you about processing in
two visual areas of the brain: the primary visual cortical area (V1),
and the middle temporal visual area (MT). To understand the behavior
of the models, the tutorial begins with a simplified version of the V1
model. Then we add some enhancements to construct a more realistic
model of V1 neural responses. Finally, we will add a second stage of
computation, corresponding to MT neurons. The application allows you
to simulate a wide variety of neurophysiology experiments.

The application runs only on MacIntosh computers with floating point
units (FPUs or math coprocessors). Most Quadra and Centris models
have FPUs. Performas do not have FPUs. PowerMacs do not have FPUs
either. If your MacIntosh does not have an FPU, you can buy a
software FPU emulator from John Neil & Associates
(email: johnneil@netcom.com, phone: 800-663-2943).

The application require at least 12M of RAM (that means you need 16M
total: 12M for the application and about 4M for the MacIntosh system).

The V1/MT model application and a (postscript) user's guide are
available on the World Wide Web at URL
http://white.stanford.edu/~heeger/v1-mt-model.html. The application
and user's guide are also available via anonymous ftp (fetch) from
white.stanford.edu (IP number 36.121.0.16) in directory /v1-mt-model/.
Installation is simple: double click on the self-extracting archive to
unpack the application.

For further information contact:

David Heeger
heeger@white.stanford.edu
(415) 723-4048

Eero Simoncelli
eero@central.cis.upenn.edu
(215) 898-0376
--------------------------------------------------------------------------