Research map / 3D vision / Human pose estimation
Learning Pose Specific Representations by Predicting Different Views
Read the paperCheck 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 |
|---|---|---|---|
| predict a different camera view instead of reconstructing the input view during pre-training (ablation) | PreView representation · NYU-CS; frozen representation with 100, 1,000, 10,000, or 43,640 labeled samples | ME: 48.0 -> 33.4; 47.2 -> 29.6; 47.3 -> 29.0; 47.1 -> 29.0 | Helped |
| use semi-supervised PreView instead of supervised training (ablation) | hand pose estimation · MV-hands (289 labeled) | ME: 26.35 -> 24.14 | Helped |
| use semi-supervised PreView instead of supervised training (ablation) | hand pose estimation · MV-hands (289 labeled) | FS80: 0.36 -> 0.39 | Helped |
| use semi-supervised PreView instead of supervised training (ablation) | hand pose estimation · MV-hands (289 labeled) | JS80: 0.67 -> 0.69 | Helped |
| increase latent representation size from 20 to 30 (ablation) | PreView · NYU validation set | ME | Helped |
| increase latent representation size above 50 (ablation) | PreView · NYU validation set | ME | No clear effect |
| add adversarial loss (ablation) | Semi-supervised PreView · NYU-CS, varying labeled sample counts | ME: 29.12 -> 29.52; 22.96 -> 23.32; 21.49 -> 20.67; 20.70 -> 20.23 | Mixed |
| add adversarial loss (ablation) | Semi-supervised PreView · MV-hands (189 labeled) | ME: 27.27 -> 27.79 | Hurt |
| use view prediction pre-training rather than Context Encoders (ablation) | PreView representation · NYU-CS; frozen representation with 100, 1,000, 10,000, or 43,640 labeled samples | ME: 53.4 -> 33.4; 53.4 -> 29.6; 53.3 -> 29.0; 53.8 -> 29.0 | Helped |
Closest papers by meaning
- Pose Prior Learner: Unsupervised Categorical Prior Learning for Pose Estimation (ICLR 2026 · similarity 0.89)
- Semi-Supervised 3D Hand-Object Poses Estimation With Interactions in Time (CVPR 2021 · similarity 0.89)
- Learning Implicit Representation for Reconstructing Articulated Objects (ICLR 2024 · similarity 0.88)
- NOPE: Novel Object Pose Estimation from a Single Image (CVPR 2024 · similarity 0.87)
- SiMHand: Mining Similar Hands for Large-Scale 3D Hand Pose Pre-training (ICLR 2025 · similarity 0.87)
- Unsupervised Learning of Category-Level 3D Pose from Object-Centric Videos (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.