buddit1 / volterra-net

VolterraNet implementation and Spherical MNIST experiment.

volterra-net

VolterraNet implementation and Spherical MNIST experiment.

Please cite the following paper if you use this code:

Banerjee, M., Chakraborty, R., Bouza, J., & Vemuri, B. C. (2020). VolterraNet: A higher order convolutional network with group equivariance for homogeneous manifolds. IEEE Transactions on Pattern Analysis and Machine Intelligence.

Requirements

• Pytorch (0.4 <= )
• Tensorflow
• s2cnn

Running Spherical MNIST

Please download the [MNIST dataset] (http://yann.lecun.com/exdb/mnist/) and place it in a folder named "mnist_data" in the project directory. Next, run the gendata.py script to generate the Spherical MNIST dataset. Now the seperate models can be trained by calling the corresponding file.