mbeyeler/motionenergy

CUDA implementation of the Motion Energy model (Simoncelli & Heeger, 1998)

MotionEnergy

This is a CUDA implementation of Simoncelli and Heeger's Motion Energy model (Simoncelli & Heeger, 1998).

The code comes with both a Python interface (in pyME) and a C/C++ interface (in cppME).

If you use this code in a scholarly publication, please cite as follows:

Beyeler, M., Dutt, N., Krichmar, J.L. (2014). Efficient Spiking Neural Network Model of Pattern Motion Selectivity in Visual Cortex Neuroinformatics 12(3):435-454, doi:10.1007/s12021-014-9220-y.

Or use the following BibTex:

@article{Beyeler2014,
	author = {M. Beyeler and N. Dutt and J. L. Krichmar},
	title = {Efficient Spiking Neural Network Model of Pattern Motion Selectivity in Visual Cortex},
	journal = {Neuroinformatics},
	year = {2014},
	volume = {12},
	number = {3},
	pages = {435--454},
	doi = {10.1007/s12021-014-9220-y}
}

Installation

1. Fork MotionEnergy by clicking on the Fork box in the top-right corner.

2. Clone the repo, where YourUsername is your actual GitHub user name:

   $ git clone https://github.com/YourUsername/MotionEnergy
   $ cd MotionEnergy
   

3. Choose whether you want to use the Python interface or the C/C++ interface.

   • Python: There is no package install yet. See the file pyME/run_dir_V1.py for an example script.
   • C++: The installation depends on your platform.

     • Linux / Mac OS X: By default, MotionEnergy gets installed to /opt/CARL/ME. You can change this by exporting an environment variable called ME_LIB_DIR:

       $ export ME_LIB_DIR=/path/to/your/preferred/dir
       

       Then compile and install:

       $ cd cppME
       $ make
       $ sudo -E make install
       

       Note the -E flag, which will cause sudo to remember the ME_LIB_DIR.

     • Windows: Simply open the solution file motion_energy.sln in Visual Studio.