Research map / 3D vision / Human pose estimation
LatentFusion: End-to-End Differentiable Reconstruction and Rendering for Unseen Object Pose Estimation
Keunhong Park, Arsalan Mousavian, Yu Xiang, Dieter Fox
Check 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 |
|---|---|---|---|
| Use ConvGRU rather than average pooling to fuse reference-view features (ablation) | LatentFusion · MOPED | ADD AUC: 57.5 -> 63.1 | Improved |
| Use ConvGRU rather than average pooling to fuse reference-view features (ablation) | LatentFusion · MOPED | ADD-S AUC: 90.1 -> 91.1 | Improved |
| Use ConvGRU rather than average pooling to fuse reference-view features (ablation) | LatentFusion · MOPED | Proj.2D AUC: 40.1 -> 44.7 | Improved |
| Increase reference views from 1 to 8 (ablation) | LatentFusion · MOPED | ADD AUC: 23.9 -> 63.1 | Improved |
| Increase reference views from 8 to 16 (ablation) | LatentFusion · MOPED | ADD AUC: 63.1 -> 62.7 | No clear effect |
| Increase reference views from 1 to 8 (ablation) | LatentFusion · MOPED | ADD-S AUC: 78.7 -> 91.1 | Improved |
| Increase reference views from 8 to 16 (ablation) | LatentFusion · MOPED | ADD-S AUC: 91.1 -> 90.8 | No clear effect |
| Increase reference views from 1 to 16 (ablation) | LatentFusion · MOPED | Proj.2D AUC: 9.5 -> 46.8 | Improved |
| Increase reference views from 16 to 32 (ablation) | LatentFusion · MOPED | Proj.2D AUC: 46.8 -> 42.6 | Hurt |
| Add the latent loss alongside the depth loss (ablation) | LatentFusion · MOPED | ADD AUC: 63.1 -> 64.2 | Mixed |
| Add the latent loss alongside the depth loss (ablation) | LatentFusion · MOPED | ADD-S AUC: 91.1 -> 91.2 | No clear effect |
| Add the latent loss alongside the depth loss (ablation) | LatentFusion · MOPED | Proj.2D AUC: 44.7 -> 43.9 | Hurt |
Later papers that compared with this method
2 later accepted papers report LatentFusion as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- Prior-free 3D Object Tracking (CVPR 2025 · better on 2 of 2 shared results)
- High-Resolution Open-Vocabulary Object 6D Pose Estimation (TPAMI 2025 · better on 1 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- pix2pose — Pix2Pose: Pixel-Wise Coordinate Regression of Objects for 6D Pose Estimation (ICCV 2019) · better on 1 of 1 shared result
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
- Pos3R: 6D Pose Estimation for Unseen Objects Made Easy (CVPR 2025 · similarity 0.89)
- One2Any: One-Reference 6D Pose Estimation for Any Object (CVPR 2025 · similarity 0.89)
- iG-6DoF: Model-free 6DoF Pose Estimation for Unseen Object via Iterative 3D Gaussian Splatting (CVPR 2025 · similarity 0.88)
- NOPE: Novel Object Pose Estimation from a Single Image (CVPR 2024 · similarity 0.88)
- 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)
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.
