Research map / 3D vision / Human pose estimation
Pose Anything Anywhere: Model-free Object Poses from Arbitrary References
Hongli XU, Jiaqi Hu, Junwen Huang, Boyang Zhong, Peter Yu, Nassir Navab, Benjamin Busam, Slobodan Ilic
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 cross-view matching on visual embeddings (ablation) | PANY · Real275 (RGB) | ADD-0.1d: 23.6 -> 25.4 | Improved |
| add geometric feature fusion to matching (ablation) | PANY · Real275 (RGB) | ADD-0.1d: 25.4 -> 27.8 | Improved |
| add cross-view matching and fusion over geometry-only (ablation) | PANY · Real275 (RGB) | AR: 56.6 -> 59.3 | Improved |
| use PANY training instead of vanilla VGGT backbone (ablation) | PANY · backbone-only baselines | AR: 43.6 -> 51.1 | Improved |
| add 8 unposed assist views at inference | PANY · LINEMOD | ADD-0.1d: 55.3 -> 87.4 | Improved |
| add 8 unposed assist views at inference | PANY · YCB-Video | ADD AUC: 92.5 -> 94.6 | Improved |
| add 8 unposed assist views at inference | PANY · LM-O | AR: 36.9 -> 42.3 | Improved |
| increase number of assist views (ablation) | PANY · LM-O | AR | Improved |
| randomly resample anchor view (ablation) | PANY · LM-O | AR | No clear effect |
| RGB-D instead of RGB input | PANY · Real275 | AR: 59.3 -> 81.8 | Improved |
| PANY vs best baseline on Toyota-Light RGB | PANY · Toyota-Light (RGB) | ADD-0.1d: 13.1 (SingRef6D) vs 10.8 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- One2Any — One2Any: One-Reference 6D Pose Estimation for Any Object (CVPR 2025) · better on 6 of 6 shared results
- SingRef6D — SingRef6D: Monocular Novel Object Pose Estimation with a Single RGB Reference (NeurIPS 2025) · better on 3 of 4 shared results
- Any6D — Any6D: Model-free 6D Pose Estimation of Novel Objects (CVPR 2025) · better on 3 of 3 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Any6D: Model-free 6D Pose Estimation of Novel Objects (CVPR 2025 · similarity 0.88)
- UNOPose: Unseen Object Pose Estimation with an Unposed RGB-D Reference Image (CVPR 2025 · similarity 0.88)
- SingRef6D: Monocular Novel Object Pose Estimation with a Single RGB Reference (NeurIPS 2025 · similarity 0.87)
- 3D-Aware Hypothesis & Verification for Generalizable Relative Object Pose Estimation (ICLR 2024 · similarity 0.86)
- RefPose: Leveraging Reference Geometric Correspondences for Accurate 6D Pose Estimation of Unseen Objects (CVPR 2025 · similarity 0.86)
- FoundationPose: Unified 6D Pose Estimation and Tracking of Novel Objects (CVPR 2024 · similarity 0.86)
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.
