Research map / Label-efficient and robust learning / Domain adaptation and generalization
Neural Pose Transfer by Spatially Adaptive Instance Normalization
Jiashun Wang, Chao Wen, Yanwei Fu, Haitao Lin, Tianyun Zou, Xiangyang Xue, Yinda Zhang
Check 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 SPAdaIN modules to the pose-transfer network (ablation) | Neural Pose Transfer · Seen-pose | PMD (× 10^-4): 8.3 to 1.1 | Improved |
| Add SPAdaIN modules to the pose-transfer network (ablation) | Neural Pose Transfer · Unseen-pose | PMD (× 10^-4): 13.7 to 9.3 | Improved |
| Add edge length regularization (ablation) | Neural Pose Transfer · Seen-pose | PMD (× 10^-4): 1.2 to 1.1 | Improved |
| Add edge length regularization (ablation) | Neural Pose Transfer · Unseen-pose | PMD (× 10^-4): 10.1 to 9.3 | Improved |
| Replace per-vertex pose features with global max pooling | Neural Pose Transfer · Seen-pose | PMD (× 10^-4): 1.1 -> 2.1 | Hurt |
| Replace per-vertex pose features with global max pooling | Neural Pose Transfer · Unseen-pose | PMD (× 10^-4): 9.3 -> 12.7 | Hurt |
| Use Neural Pose Transfer instead of DT with 5 control points | Pose transfer · Seen-pose | PMD (× 10^-4): 7.3 -> 1.1 | Improved |
| Use Neural Pose Transfer instead of DT with 13 control points | Pose transfer · Seen-pose | PMD (× 10^-4): 7.7 -> 1.1 | Improved |
| Use Neural Pose Transfer instead of DT with 5 control points | Pose transfer · Unseen-pose | PMD (× 10^-4): 7.2 -> 9.3 | Hurt |
| Use Neural Pose Transfer instead of DT with 13 control points | Pose transfer · Unseen-pose | PMD (× 10^-4): 6.7 -> 9.3 | Hurt |
| Use Neural Pose Transfer instead of the skeleton-based approach | Pose transfer · Seen-pose | PMD (× 10^-4): 27.4 -> 1.1 | Improved |
| Use Neural Pose Transfer instead of the skeleton-based approach | Pose transfer · Unseen-pose | PMD (× 10^-4): 31.1 -> 9.3 | Improved |
Closest papers by meaning
- 3D Pose Transfer with Correspondence Learning and Mesh Refinement (NeurIPS 2021 · similarity 0.91)
- Unsupervised 3D Pose Transfer With Cross Consistency and Dual Reconstruction (TPAMI 2023 · similarity 0.88)
- Neural Pose Representation Learning for Generating and Transferring Non-Rigid Object Poses (NeurIPS 2024 · similarity 0.87)
- Weakly-supervised 3D Pose Transfer with Keypoints (ICCV 2023 · similarity 0.86)
- Towards Robust 3D Pose Transfer with Adversarial Learning (CVPR 2024 · similarity 0.85)
- REACTO: Reconstructing Articulated Objects from a Single Video (CVPR 2024 · similarity 0.84)
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.
