Research map / 3D vision / NeRF and radiance fields
Canonical Fields: Self-Supervised Learning of Pose-Canonicalized Neural Fields
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 |
|---|---|---|---|
| Apply CaFi-Net directly to neural fields rather than using point-cloud ConDor | Canonicalization on 13 ShapeNet categories · 13 ShapeNet categories (avg.) | CC: 1.72 -> 1.45 | Helped |
| Apply CaFi-Net directly to neural fields rather than using point-cloud ConDor | Canonicalization on 13 ShapeNet categories · 13 ShapeNet categories (avg.) | GEC: 1.81 -> 1.67 | Helped |
| Apply CaFi-Net directly to neural fields rather than using point-cloud ConDor | Canonicalization on 13 ShapeNet categories · 13 ShapeNet categories (avg.) | IC: 1.15 -> 1.34 | No clear effect |
| Use density gradients instead of xyz locations as the input signal (ablation) | CaFi-Net · 4 categories (avg.; bench, cellphone, chair, plane) | GEC: 1.95 -> 1.57 | Mixed |
| Use density gradients instead of xyz locations as the input signal (ablation) | CaFi-Net · 4 categories (avg.; bench, cellphone, chair, plane) | IC: 1.73 -> 1.26 | Helped |
| Use density gradients instead of xyz locations as the input signal (ablation) | CaFi-Net · 4 categories (avg.; bench, cellphone, chair, plane) | CC: 1.72 -> 1.35 | Helped |
| Use density weighting instead of local average density weighting for equivariant features (ablation) | CaFi-Net · 4 categories (avg.; bench, cellphone, chair, plane) | GEC: 1.73 -> 1.57 | Helped |
| Use density weighting instead of local average density weighting for equivariant features (ablation) | CaFi-Net · 4 categories (avg.; bench, cellphone, chair, plane) | IC: 1.85 -> 1.26 | Helped |
| Use density weighting instead of local average density weighting for equivariant features (ablation) | CaFi-Net · 4 categories (avg.; bench, cellphone, chair, plane) | CC: 1.49 -> 1.35 | Helped |
| Add Siamese training to align instances of the same category (ablation) | CaFi-Net · 4 categories (avg.; bench, cellphone, chair, plane) | GEC: 1.86 -> 1.57 | Helped |
| Add Siamese training to align instances of the same category (ablation) | CaFi-Net · 4 categories (avg.; bench, cellphone, chair, plane) | IC: 1.48 -> 1.26 | Helped |
| Add Siamese training to align instances of the same category (ablation) | CaFi-Net · 4 categories (avg.; bench, cellphone, chair, plane) | CC: 1.62 -> 1.35 | Helped |
Closest papers by meaning
- One-shot 3D Object Canonicalization based on Geometric and Semantic Consistency (CVPR 2025 · similarity 0.87)
- Joint Optimization of Neural Radiance Fields and Continuous Camera Motion from a Monocular Video (CVPR 2025 · similarity 0.86)
- ImageNet3D: Towards General-Purpose Object-Level 3D Understanding (NeurIPS 2024 · similarity 0.85)
- Mask-Based Modeling for Neural Radiance Fields (ICLR 2024 · similarity 0.85)
- CF-NeRF: Camera Parameter Free Neural Radiance Fields with Incremental Learning (AAAI 2024 · similarity 0.85)
- ICON: Incremental CONfidence for Joint Pose and Radiance Field Optimization (CVPR 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.