We propose a semantic segmentation model that exploits rotation and reflection symmetries. We demonstrate significant gains in sample efficiency due to increased weight sharing, as well as improvements in robustness to symmetry transformations. The group equivariant CNN framework is extended for segmentation by introducing a new equivariant (G->Z2)-convolution that transforms feature maps on a group to planar feature maps. Also, equivariant transposed convolution is formulated for up-sampling in an encoder-decoder network. To demonstrate improvements in sample efficiency we evaluate on multiple data regimes of a rotation-equivariant segmentation task: cancer metastases detection in histopathology images. We further show the effectiveness of exploiting more symmetries by varying the size of the group.
@article{arxiv.1807.00583,
title = {Sample Efficient Semantic Segmentation using Rotation Equivariant Convolutional Networks},
author = {Jasper Linmans and Jim Winkens and Bastiaan S. Veeling and Taco S. Cohen and Max Welling},
journal= {arXiv preprint arXiv:1807.00583},
year = {2018}
}
Comments
Presented at the ICML workshop: Towards learning with limited labels: Equivariance, Invariance, and Beyond, 2018