Research map / Architectures and efficiency / CNN design
Scalable and Equivariant Spherical CNNs by Discrete-Continuous (DISCO) Convolutions
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 |
|---|---|---|---|
| Use DISCO-Axisymmetric instead of planar convolutions | Spherical MNIST classifier · Spherical MNIST R/R | accuracy: 53.9 -> 98.7 | Helped |
| Use DISCO-Axisymmetric instead of planar convolutions | Spherical MNIST classifier · Spherical MNIST NR/R | accuracy: 14.3 -> 98.6 | Helped |
| Add random mirroring augmentation (ablation) | DISCO-Directional · 2D3DS semantic segmentation | mIoU: 45.2 -> 45.7 | Helped |
| Use directional rather than axisymmetric filters (ablation) | DISCO spherical CNN · 2D3DS semantic segmentation | mIoU: 39.7 -> 45.2 | Helped |
| Add copy-and-paste augmentation (ablation) | DISCO-Separable · Omni-SYNTHIA semantic segmentation | mIoU: 48.3 -> 49.2 | Helped |
| Use DISCO-Directional rather than Planar UNet | Pano3D depth estimation · Pano3D | wRMSE: 0.4520 -> 0.5063 | Hurt |
| Use DISCO-Directional rather than Planar UNet | Pano3D depth estimation · Pano3D | δico1.05: 36.68% -> 38.32% | Helped |
Closest papers by meaning
- Efficient Generalized Spherical CNNs (ICLR 2021 · similarity 0.93)
- Equivariance versus Augmentation for Spherical Images (ICML 2022 · similarity 0.92)
- DeepSphere: a graph-based spherical CNN (ICLR 2020 · similarity 0.91)
- Spin-Weighted Spherical CNNs (NeurIPS 2020 · similarity 0.91)
- Unified Spherical Frontend: Learning Rotation-Equivariant Representations of Spherical Images from Any Camera (CVPR 2026 · similarity 0.90)
- Rotation Equivariant Graph Convolutional Network for Spherical Image Classification (CVPR 2020 · similarity 0.90)
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.