atlas

Research map / Architectures and efficiency / CNN design

FILTRA: Rethinking Steerable CNN by Filter Transform

ICML 2021 · Architectures and efficiency › CNN design · approach: Steerable rotation-equivariant CNNs · 0 citations (OpenAlex)

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Use FILTRA instead of R2Convclassification CNN · MNIST, SO2 augmentation, C8accuracy: 98.8 -> 98.9Helped
Use FILTRA instead of R2Convclassification CNN · KMNIST, SO2 augmentation, C8accuracy: 97.3 -> 97.1Hurt
Use FILTRA instead of R2Convclassification CNN · KMNIST, O2 augmentation, D8accuracy: 96.8 -> 97.0Helped
Use FILTRA instead of R2Convclassification CNN · FashionMNISTaccuracy: 90.5 -> 90.5; 90.8 -> 90.8No clear effect
Use FILTRA instead of R2Convclassification CNN · EMNISTaccuracy: 76.7 -> 77.1; 80.1 -> 80.5Helped
Use FILTRA instead of R2ConvWideResNet · CIFAR10, SO2 augmentation, C4accuracy: 93.6 -> 93.4Hurt
Use FILTRA instead of R2ConvWideResNet · CIFAR10, O2 augmentation, D4accuracy: 92.7 -> 92.8Helped
Use FILTRA instead of R2Convcharacter-direction regressor · MNIST, SO2 augmentation, C8angle error (deg): 4.8 -> 3.3Helped
Use FILTRA instead of R2Convcharacter-direction regressor · EMNIST, O2 augmentation, D8angle error (deg): 29.2 -> 24.9Helped

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy