rgc-natstim-model

Description

Companion repository to publication "A chromatic feature detector in the retina signals visual context changes".

We trained CNN models (digital twins) of mouse retinal processing of naturalistic stimuli. We then used the models to analyse neuronal stimulus selectivities in-silico and found a selectivity for chromatic contrast in a type of contrast-suppressed retinal ganglion cell (RGC). Based on this feature, we proposed a role in detecting visual context changes for this RGC type.

This repository contains the code to reproduce the analyses and figures presented in the paper.

How to use this repository

1. Clone this repository, navigate to its directory, and install it via
   pip install . Also install packages listed in the requirements.txt file.
2. Download the data and model files from G-Node . Update the base_directory in rgc_natstim_model/constants/paths.py to point to the respective directory on your machine.

Reproducing figures

Run the notebooks.

Training models

Run the model_training notebook. In order to generate MEIs, run the mei_generation notebook. Model and MEI optimization functionality is implemented in open-retina.