Skip to content

Latest commit

 

History

History
10 lines (6 loc) · 262 Bytes

README.md

File metadata and controls

10 lines (6 loc) · 262 Bytes

groupcnn

Group equivariant CNN in pytorch.

This is a partial implementation of neural modules from the paper

'Group Equivariant Convolutional Networks'

by T. Cohen and M. Welling (2016).

So far only rotation equivariant convolutions (P4) are implemented.