Research map / Architectures and efficiency / CNN design
Integrating Symmetry into Differentiable Planning with Steerable 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 |
|---|---|---|---|
| add D4-equivariant steerable convolutions to VIN | VIN · 2D navigation 15×15 | test success rate (%): 66.97 -> 98.99 | Helped |
| add D4-equivariant steerable convolutions to VIN | VIN · 2D navigation 28×28 | test success rate (%): 67.57 -> 98.14 | Helped |
| add D4-equivariant steerable convolutions to VIN | VIN · 2D navigation 50×50 | test success rate (%): 57.92 -> 86.20 | Helped |
| add D4-equivariant steerable convolutions to VIN | VIN · Visual navigation | test success rate (%): 50.83 -> 95.50 | Helped |
| add D4-equivariant steerable convolutions to VIN | VIN · 2-DOF C-space manipulation 18×18 | test success rate (%): 77.82 -> 99.98 | Helped |
| add D4-equivariant steerable convolutions to VIN | VIN · 2-DOF C-space manipulation 36×36 | test success rate (%): 84.32 -> 99.36 | Helped |
| add D4-equivariant steerable convolutions to VIN | VIN · Workspace manipulation | test success rate (%): 80.44 -> 91.10 | Helped |
| replace standard convolutions with steerable convolutions | ConvGPPN · 2D navigation 15×15 | test success rate (%): 99.75 -> 99.98 | Helped |
| replace standard convolutions with steerable convolutions | ConvGPPN · 2D navigation 28×28 | test success rate (%): 99.09 -> 99.86 | Helped |
| replace standard convolutions with steerable convolutions | ConvGPPN · 2D navigation 50×50 | test success rate (%): 97.21 -> 99.49 | Helped |
| replace standard convolutions with steerable convolutions | ConvGPPN · Visual navigation | test success rate (%): 98.55 -> 99.78 | Helped |
| replace standard convolutions with steerable convolutions | ConvGPPN · 2-DOF C-space manipulation 18×18 | test success rate (%): 99.98 -> 100.00 | Helped |
Closest papers by meaning
- Scaling Value Iteration Networks to 5000 Layers for Extreme Long-Term Planning (ICML 2025 · similarity 0.86)
- Versatile Navigation Under Partial Observability via Value-guided Diffusion Policy (CVPR 2024 · similarity 0.86)
- SiT: Symmetry-invariant Transformers for Generalisation in Reinforcement Learning (ICML 2024 · similarity 0.85)
- SPACE: Unifying Symmetric and Asymmetric Routing Problems for Generalist Neural Solver (NeurIPS 2026 · similarity 0.84)
- Seeing through Uncertainty: Robust Task-Oriented Optimization in Visual Navigation (NeurIPS 2025 · similarity 0.84)
- Physics-informed Temporal Difference Metric Learning for Robot Motion Planning (ICLR 2025 · similarity 0.84)
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.