A radially symmetric convolution kernel, SRE-Conv, improves rotated-image classification accuracy on all 16 MedMNISTv2 tasks while using fewer parameters than standard CNNs.
Our results demonstrate that our SRE-Conv layers with equivariant kernels improve accu- racy and reliability across a wide range of biomedical imaging applications
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
SRE-Conv: Symmetric Rotation Equivariant Convolution for Biomedical Image Classification
A radially symmetric convolution kernel, SRE-Conv, improves rotated-image classification accuracy on all 16 MedMNISTv2 tasks while using fewer parameters than standard CNNs.