Rotation Equivariance and Invariance in Convolutional Neural Networks — mentioned by improbable22

Fri, 21 Sep 2018 18:45:16 GMT

There is work on rotationally invariant networks, e.g. for identifying galaxies, or cells under a microscope. For example: https://arxiv.org/abs/1612.04642 https://arxiv.org/abs/1805.12301 I haven't looked closely enough to be sure if they literally had complex activations, but this seems like an obvious use. Maybe they would have, if only tensorflow made it easy.

Paper and discussion history. A mention is not an endorsement.

Rotation Equivariance and Invariance in Convolutional Neural Networks · HackerPapers discussions

Rotation Equivariance and Invariance in Convolutional Neural Networks — mentioned by improbable22