The paper proves that all O(d)-equivariant convolution kernels are sums of radial functions times products of coordinate components and Kronecker deltas, and implements this in standard CNNs for biomedical tasks.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Moment kernels: a simple and scalable approach for equivariance to rotations and reflections in deep convolutional networks
The paper proves that all O(d)-equivariant convolution kernels are sums of radial functions times products of coordinate components and Kronecker deltas, and implements this in standard CNNs for biomedical tasks.