Research map / 3D vision / Human pose estimation
Weakly-supervised 3D Pose Transfer with Keypoints
Jinnan Chen, Chen Li, Gim Hee Lee
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 |
|---|---|---|---|
| add cycle reconstruction loss (ablation) | keypoint-based pose transfer · NPT dataset | PMD (1e-4): 6.33 -> 1.47 | Improved |
| add GMM-based pseudo skinning weights loss (ablation) | keypoint-based pose transfer · NPT dataset | PMD (1e-4): 6.48 -> 1.47 | Improved |
| add mesh refinement network (ablation) | keypoint-based pose transfer · NPT dataset | PMD (1e-4): 3.28 -> 1.47 | Improved |
| add self reconstruction loss (ablation) | keypoint-based pose transfer · FAUST dataset | PMD (1e-4): 18.1 -> 11.7 | Improved |
| add edge length loss (ablation) | keypoint-based pose transfer · FAUST dataset | PMD (1e-4): 14.9 -> 11.7 | Improved |
| use weak keypoint supervision instead of fully-supervised ground-truth meshes | pose transfer · NPT dataset | PMD (1e-4): 1.23 -> 1.47 | Mixed |
| use keypoint-based IK pose transfer instead of implicit unsupervised latent codes | pose transfer on FAUST · FAUST dataset | PMD (1e-4): 22.31 -> 11.70 | Improved |
| use weakly-supervised keypoint method for cross-dataset generalization | pose transfer on Mixamo · Mixamo (cross-dataset) | PMD (1e-3): 23.5 -> 15.0 | Improved |
Closest papers by meaning
- MAPConNet: Self-supervised 3D Pose Transfer with Mesh and Point Contrastive Learning (ICCV 2023 · similarity 0.90)
- Unsupervised 3D Pose Transfer With Cross Consistency and Dual Reconstruction (TPAMI 2023 · similarity 0.89)
- Self-Supervised 3D Human Pose Estimation via Part Guided Novel Image Synthesis (CVPR 2020 · similarity 0.88)
- 3D Pose Transfer with Correspondence Learning and Mesh Refinement (NeurIPS 2021 · similarity 0.88)
- Towards 3D Human Pose Estimation in the Wild: A Weakly-Supervised Approach (ICCV 2017 · similarity 0.88)
- Neural Pose Representation Learning for Generating and Transferring Non-Rigid Object Poses (NeurIPS 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.
