Research map / Label-efficient and robust learning / Domain adaptation and generalization
UDA-COPE: Unsupervised Domain Adaptation for Category-Level Object Pose Estimation
Taeyeop Lee, Byeong-Uk Lee, Inkyu Shin, Jaesung Choe, Ukcheol Shin, In So Kweon, Kuk-Jin Yoon
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 |
|---|---|---|---|
| Train with unlabeled REAL data using the proposed pseudo-label adaptation | UDA-COPE RGB · REAL275 | 5°5cm: 18.3 -> 27.0 (+8.7) | Improved |
| Train with unlabeled REAL data using the proposed pseudo-label adaptation | UDA-COPE RGB · REAL275 | 3D50: 79.9 -> 82.0 (+2.1) | Improved |
| Add unlabeled REAL training data with rendered-depth self-supervision | CPS++ RGB · REAL275 | 5°5cm: 25.8 -> 25.2 (-0.6) | Hurt |
| Add momentum update to naive pseudo-label training (ablation) | UDA-COPE · REAL275 | 5°5cm: 28.2 -> 27.8 | Hurt |
| Add all-modality loss to pseudo-label training (ablation) | UDA-COPE RGB-D · REAL275 | 5°5cm: 28.2 -> 29.2 | Improved |
| Use xMUDA consistency instead of all-modality loss (ablation) | UDA-COPE RGB-D · REAL275 | 5°5cm: 28.2 -> 30.3 | Improved |
| Add bidirectional point filtering to pseudo-label training (ablation) | UDA-COPE RGB-D · REAL275 | 5°5cm: 29.2 -> 32.9 | Improved |
| Add teacher self-supervised learning using filtered depth points (ablation) | UDA-COPE RGB-D · REAL275 | 5°5cm: 32.9 -> 34.8 (+6.9) | Improved |
| Use bidirectional point filtering instead of confidence, entropy, or prediction-agreement filtering (ablation) | UDA-COPE · REAL275 | 5°5cm: RGB: 25.0 -> 26.1; Depth: 26.1 -> 27.3; RGB-D: 30.3 -> 32.9 | Mixed |
Closest papers by meaning
- Source-Free and Image-Only Unsupervised Domain Adaptation for Category Level Object Pose Estimation (ICLR 2024 · similarity 0.92)
- TTA-COPE: Test-Time Adaptation for Category-Level Object Pose Estimation (CVPR 2023 · similarity 0.91)
- U-COPE: Taking a Further Step to Universal 9D Category-level Object Pose Estimation (ECCV 2024 · similarity 0.90)
- NOPE: Novel Object Pose Estimation from a Single Image (CVPR 2024 · similarity 0.88)
- Unsupervised Learning of Category-Level 3D Pose from Object-Centric Videos (CVPR 2024 · similarity 0.87)
- Co-op: Correspondence-based Novel 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.
