Deep Learning Models of the Retinal Response to Natural Scenes

Deep retina is a project to test to what degree artificial neural networks can predict retinal ganglion cell responses to natural stimuli.

Please see our NIPS paper for more details.

Note that deepretina requires python 3.5 or higher.

Usage

To install the dependencies, run pip install -r requirements.txt. Scripts for training models are located in the scripts folder (e.g. see scripts/fit_models.py). Model definitions are in deepretina/models.py