Research map / Architectures and efficiency / CNN design
Steerable CNNs
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 changed | Where | Measured | Effect |
|---|---|---|---|
| replace standard convolutions with steerable convolutions | ResNet-style CNN · CIFAR10+ (50k labels) | test error: 4.17 -> 3.65 | Helped |
| replace standard convolutions with steerable convolutions | ResNet-style CNN · CIFAR100+ (50k labels) | test error: 19.25 -> 18.82 | Helped |
| use mixed quotient and irreducible capsules instead of standard ResNet features | Steerable CNN · CIFAR10 (2k labels) | test error: 30% -> 24.48% | Helped |
| use mixed quotient and irreducible capsules instead of only regular capsules | Steerable CNN · CIFAR10 (2k labels) | test error: 26.75% -> 24.48% | Helped |
| restrict the network to a single capsule type other than regular capsules (ablation) | Steerable CNN · CIFAR10 (2k labels) | test error: roughly 30%-40% | Hurt |
| use regular-representation capsules rather than standard CNN features (ablation) | Steerable CNN · CIFAR10 (2k labels) | test error: 30% -> 26.75% | Helped |
Closest papers by meaning
- General E(2)-Equivariant Steerable CNNs (NeurIPS 2019 · similarity 0.91)
- Learning Steerable Filters for Rotation Equivariant CNNs (CVPR 2018 · similarity 0.91)
- Steerable Transformers for Volumetric Data (ICML 2025 · similarity 0.89)
- A Probabilistic Approach to Learning the Degree of Equivariance in Steerable CNNs (ICML 2024 · similarity 0.89)
- Clifford-Steerable Convolutional Neural Networks (ICML 2024 · similarity 0.88)
- PICNN: A Pathway towards Interpretable Convolutional Neural Networks (AAAI 2024 · similarity 0.85)
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.