Research map / 3D vision / Human pose estimation
View-Invariant Probabilistic Embedding for Human Pose
Jennifer J. Sun, Jiaping Zhao, Liang-Chieh Chen, Florian Schroff, Hartwig Adam, Ting Liu
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 |
|---|---|---|---|
| use probabilistic instead of point embedding (ablation) | VIPE · H3.6M | Hit@1 (%): 75.4% -> 76.2% | Improved |
| use probabilistic instead of point embedding (ablation) | VIPE · 3DHP | Hit@1 (%): 19.7% -> 20.0% | Improved |
| use probabilistic instead of point embedding, with camera augmentation (ablation) | VIPE · H3.6M | Hit@1 (%): 73.8% -> 73.7% | No clear effect |
| use probabilistic instead of point embedding, with camera augmentation (ablation) | VIPE · 3DHP | Hit@1 (%): 26.1% -> 26.5% | Improved |
| add camera augmentation (ablation) | Pr-VIPE · 3DHP (all cameras) | Hit@1 (%): 19.9 -> 26.4 | Improved |
| add camera augmentation (ablation) | Pr-VIPE · H3.6M | Hit@1 (%): 76.2 -> 73.7 | Hurt |
| add camera augmentation (ablation) | L2-VIPE · 3DHP (all cameras) | Hit@1 (%): 18.7 -> 23.7 | Improved |
| increase embedding dimension from 16 to 32 (ablation) | Pr-VIPE (w/ aug.) · H3.6M | Hit@1 (%): 73.7% -> 75.5% | Improved |
| increase number of samples K from 10 to 20 (ablation) | Pr-VIPE · H3.6M | Hit@1: 0.744 -> 0.762 | Improved |
| increase number of samples K from 20 to 30 (ablation) | Pr-VIPE · H3.6M | Hit@1: 0.762 -> 0.755 | No clear effect |
| vary triplet margin beta from 1.25 to 3 (ablation) | Pr-VIPE · H3.6M | Hit@1: 0.758 -> 0.760 | No clear effect |
| widen augmentation elevation/roll range from +-15 to +-45 degrees (ablation) | Pr-VIPE · 3DHP (chest) | Hit@1: 0.289 -> 0.273 | Hurt |
Closest papers by meaning
- View-Invariant, Occlusion-Robust Probabilistic Embedding for Human Pose (IJCV 2021 · similarity 0.94)
- PIEs: Pose Invariant Embeddings (CVPR 2019 · similarity 0.88)
- Multiview-Consistent Semi-Supervised Learning for 3D Human Pose Estimation (CVPR 2020 · similarity 0.87)
- Thin-Slicing for Pose: Learning to Understand Pose Without Explicit Pose Estimation (CVPR 2016 · similarity 0.87)
- Dual Pose-invariant Embeddings: Learning Category and Object-specific Discriminative Representations for Recognition and Retrieval (CVPR 2024 · similarity 0.87)
- ProbPose: A Probabilistic Approach to 2D Human Pose Estimation (CVPR 2025 · 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.
