Research map / 3D vision / Human pose estimation
FS-Net: Fast Shape-Based Network for Category-Level 6D Object Pose Estimation With Decoupled Rotation Mechanism
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 |
|---|---|---|---|
| add decoupled rotation to G2L (ablation) | G2L · NOCS-REAL | 10°10 cm: 31.0% -> 47.81% | Helped |
| add decoupled rotation to G2L (ablation) | G2L · NOCS-REAL | IoU_50: 94.65% -> 96.21% | Helped |
| add observed points reconstruction (ablation) | 3DGC shape network with DEF and DR (Med3) · NOCS-REAL | 10°10 cm: 61.1% -> 65.9% | Helped |
| add observed points reconstruction (ablation) | 3DGC shape network with DEF and no DR (Med1) · NOCS-REAL | 10°10 cm: 46.4% -> 46.8% | No clear effect |
| add decoupled rotation (ablation) | 3DGC shape network with DEF and OPR · NOCS-REAL | 10°10 cm: 46.8% -> 65.9% | Helped |
| remove online 3D deformation (ablation) | FS-Net with OPR and DR · NOCS-REAL | 10°10 cm: 65.9% -> 58.2% | Hurt |
| remove online 3D deformation (ablation) | FS-Net with OPR and DR · NOCS-REAL | IoU_50: 98.04% -> 97.30% | Hurt |
| reconstruct the complete object model instead of observed points (ablation) | FS-Net autoencoder · NOCS-REAL | 10°10 cm: 65.9% -> 58.0% | Hurt |
| use shape-based FS-Net instead of CASS | category-level pose estimation · NOCS-REAL | 10°10 cm: 58.3% -> 64.6% | Helped |
| use shape-based FS-Net instead of NOCS | category-level pose estimation · NOCS-REAL | IoU_50: 80.5% -> 92.2% | Helped |
| use FS-Net instead of PVN3D | instance-level pose estimation · LINEMOD | ADD-(S): 99.4% -> 97.6% | Hurt |
| use FS-Net instead of PVN3D | instance-level pose estimation · LINEMOD | speed (FPS): 5 -> 20 | Helped |
Closest papers by meaning
- SAR-Net: Shape Alignment and Recovery Network for Category-Level 6D Object Pose and Size Estimation (CVPR 2022 · similarity 0.90)
- One2Any: One-Reference 6D Pose Estimation for Any Object (CVPR 2025 · similarity 0.89)
- CAP-Net: A Unified Network for 6D Pose and Size Estimation of Categorical Articulated Parts from a Single RGB-D Image (CVPR 2025 · similarity 0.88)
- SchemaPose: RGB-Based Category-Level Pose Estimation with Parametric Category Schema (NeurIPS 2026 · similarity 0.87)
- CatFormer: Category-Level 6D Object Pose Estimation with Transformer (AAAI 2024 · similarity 0.87)
- GIVEPose: Gradual Intra-class Variation Elimination for RGB-based Category-Level Object Pose Estimation (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.