Research map / Detection and segmentation / CNN features
Reflection and Rotation Symmetry Detection via Equivariant Learning
Check 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 equivariant convolutions instead of standard convolutions (ablation) | EquiSym-ref · DENDI reflection | F1 score: 55.1 to 63.1 | Helped |
| Use equivariant convolutions instead of standard convolutions (ablation) | EquiSym-rot · DENDI rotation | F1 score: 17.7 to 21.2 | Helped |
| Add reflection-orientation auxiliary classification (ablation) | EquiSym-ref · DENDI reflection | F1 score: 63.1 to 64.5 | Helped |
| Add rotation-fold auxiliary classification (ablation) | EquiSym-rot · DENDI rotation | F1 score: 21.2 to 22.5 | Helped |
| Add auxiliary classification during joint training (ablation) | EquiSym-joint · DENDI joint reflection | F1 score: 62.2 to 58.7 | Hurt |
| Add auxiliary classification during joint training (ablation) | EquiSym-joint · DENDI joint rotation | F1 score: 22.1 to 22.5 | Helped |
| Train EquiSym-ref on real images rather than real and synthetic images (ablation) | EquiSym-ref · SDRW reflection | F1 score: 67.4 and 67.1 | No clear effect |
| Train EquiSym-ref on real images rather than real and synthetic images (ablation) | EquiSym-ref · LDRS reflection | F1 score: 39.4 to 40.9 | Helped |
| Use EquiSym-ref instead of PMCNet | Reflection symmetry detection · DENDI reflection | F1 score: 52.0 to 64.5 | Helped |
| Use EquiSym-rot instead of SymResNet | Rotation symmetry detection · DENDI rotation | F1 score: 11.9 to 22.5 | Helped |
| Use EquiSym-ref instead of PMCNet | Reflection symmetry detection · Mixed reflection | F1 score: 61.2 to 71.4 | Helped |
Closest papers by meaning
- R2Det: Exploring Relaxed Rotation Equivariance in 2D Object Detection (ICLR 2025 · similarity 0.91)
- Relaxed Rotational Equivariance via G-Biases in Vision (AAAI 2025 · similarity 0.89)
- Probing Equivariance and Symmetry Breaking in Convolutional Networks (NeurIPS 2025 · similarity 0.88)
- Symmetry Strikes Back: From Single-Image Symmetry Detection to 3D Generation (CVPR 2025 · similarity 0.88)
- Symmetry Discovery Beyond Affine Transformations (NeurIPS 2024 · similarity 0.88)
- Leveraging 3D Geometric Priors in 2D Rotation Symmetry Detection (CVPR 2025 · similarity 0.87)
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.