Research map / 3D vision / Human pose estimation
Environment-Agnostic Pose: Generating Environment-independent Object Representations for 6D Pose Estimation
Shaobo Zhang, Yuhang Huang, Wanqing Zhao, Wei Zhao, Ziyu Guan, Jinye Peng
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 |
|---|---|---|---|
| generate environment-independent object representations with a diffusion model instead of estimating pose directly from the image | EA6D · Linemod-Occluded | ADD(-S) average recall: 79.6 -> 87.8 | Improved |
| train only on synthetic data with EA6D instead of Self6D++ | EA6D · Linemod-Occluded | ADD(-S) average recall: 52.9 -> 87.8 | Improved |
| test EA6D on an unseen scene dataset (HomebrewedDB) | EA6D · HomebrewedDB | mean ADD(-S): 89.4 -> 94.7 | Improved |
| test EA6D on DiverseScenes with strong background and lighting shifts | EA6D · DiverseScenes | mean ADD(-S): 72.3 -> 83.8 | Improved |
| start reverse process from input latent plus noise (EDDM) instead of pure noise (DDPM) (ablation) | EA6D · Linemod (1 diffusion step) | ADD(-S) mean: 0 -> 96.3 | Improved |
| add environment decoupling gradient (EDDM) versus plain decoupled diffusion model (ablation) | EA6D · Linemod (1 diffusion step) | ADD(-S) mean: 93.1 -> 96.3 | Improved |
| use VAE latent codes of pure object images as target representation (ablation) | EA6D · Linemod (1 diffusion step) | ADD(-S) mean: 68.4 -> 96.3 | Improved |
| feed latent codes to pose decoder instead of decoding to pure images first (EA6D*) (ablation) | EA6D · Linemod (1 diffusion step) | ADD(-S) mean: 76.7 -> 96.3 | Improved |
| increase diffusion steps from 1 to 1000 (ablation) | EA6D · Linemod | ADD(-S) mean: 96.3 -> 96.5 | No clear effect |
| EA6D versus SO-Pose on Linemod | EA6D · Linemod | ADD(-S) average recall: 96.0 -> 96.3 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 on most shared results.
- 6D-diff — 6D-Diff: A Keypoint Diffusion Framework for 6D Object Pose Estimation (CVPR 2024) · better on 2 of 2 shared results
- TexPose — TexPose: Neural Texture Learning for Self-Supervised 6D Object Pose Estimation (CVPR 2023) · better on 2 of 2 shared results
- Self6D++ — Self6D: Self-Supervised Monocular 6D Object Pose Estimation (ECCV 2020) · better on 2 of 2 shared results
- SO-Pose — SO-Pose: Exploiting Self-Occlusion for Direct 6D Pose Estimation (ICCV 2021) · 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
- Confronting Ambiguity in 6D Object Pose Estimation via Score-Based Diffusion on SE(3) (CVPR 2024 · similarity 0.89)
- 6D-Diff: A Keypoint Diffusion Framework for 6D Object Pose Estimation (CVPR 2024 · similarity 0.89)
- Any6D: Model-free 6D Pose Estimation of Novel Objects (CVPR 2025 · similarity 0.87)
- iG-6DoF: Model-free 6DoF Pose Estimation for Unseen Object via Iterative 3D Gaussian Splatting (CVPR 2025 · similarity 0.87)
- Object Pose Estimation via the Aggregation of Diffusion Features (CVPR 2024 · similarity 0.87)
- FoundationPose: Unified 6D Pose Estimation and Tracking of Novel Objects (CVPR 2024 · 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.
