This folder contains the code to reproduce the models and results from the following paper:
Alexander S. Ecker, Fabian H. Sinz, Emmanouil Froudarakis, Paul G. Fahey, Santiago A. Cadena, Edgar Y. Walker, Erick Cobos, Jacob Reimer, Andreas S. Tolias, Matthias Bethge: A rotation-equivariant convolutional neural network model of primary visual cortex. International Conference on Learning Representations (ICLR 2019). https://openreview.net/forum?id=H1fU8iAqKX.
Download the checkpoints and data files (two files: checkpoints.tar.gz.aa and checkpoints.tar.gz.ab) from Github and extract them into the folder analysis/iclr2019/checkpoints using the following command:
cat checkpoints.tar.gz.* | tar xvfz -
If you're just interested in reproducing the various mdoels and baselines described in the paper, check out the Jupyter notebook models. It contains all the relevant models ready-to-use. All code necessary to construct, train and evaluate the models is contained in the module cnn_sys_ident.architectures.
The Jupyter notebooks in this folder contain the code we used to analyze the results and generate the figrues in the paper. Note, however, that they won't run out-of-the-box, as they depend on a MySQL database and a data management tool (DataJoint) that we use to keep track of our experiments. Therefore, reproducing the experiments is a bit mode involved than simply running a script, as it requires setting up DataJoint and the MySQL server. If you're interested in going that route, don't hesitate to touch base; we're happy to help.