Research map / 3D vision / Human pose estimation
Unsupervised Cross-Modal Alignment for Multi-Person 3D Pose Estimation
Jogendra Nath Kundu, Ambareesh Revanur, Govind Vitthal Waghmare, Rahul Mysore Venkatesh, R. Venkatesh Babu
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 reconstruction loss L_recon (ablation) | distillation-only unsupervised model · MuPoTS-3D (MuCo-3DHP training) | 3DPCK: 53.3 -> 57.6 | Improved |
| add cross-modal adaptation loss L_adapt (ablation) | unsupervised model with L_distl and L_recon · MuPoTS-3D (MuCo-3DHP training) | 3DPCK: 57.6 -> 61.9 | Improved |
| add self-supervised equivariance loss L_ss (ablation) | unsupervised model · MuPoTS-3D (MuCo-3DHP training) | 3DPCK: 61.9 -> 64.2 | Improved |
| use MoCap poses instead of artificial poses (ablation) | Ours-Us unsupervised · MuPoTS-3D (MuCo-3DHP training) | 3DPCK: 64.2 -> 66.1 | Improved |
| use MoCap poses instead of artificial poses (ablation) | weakly supervised model · MuPoTS-3D (MuCo-3DHP training) | 3DPCK: 66.4 -> 67.9 | No clear effect |
| add paired 2D supervision | unsupervised model · MuPoTS-3D (MuCo-3DHP training) | 3DPCK: 66.1 -> 67.9 | Improved |
| add 3D supervision and 3D reconstruction loss | weakly supervised model · MuPoTS-3D (MuCo-3DHP training) | 3DPCK: 67.9 -> 75.8 | Improved |
| add synthetic data D_syn (ablation) | fully supervised model · MuPoTS-3D (MuCo-3DHP training) | 3DPCK: 71.1 -> 75.8 | Improved |
| add Stage-2 single-person refinement network | Ours-Fs · MuPoTS-3D (MuCo-3DHP training) | 3DPCK: 75.8 -> 76.9 | Improved |
| add Stage-2 single-person refinement network | Ours-Fs · MuPoTS-3D (MuCo-3DHP training) | fps: 21.2 -> 16.6 | Hurt |
| train student with full loss instead of distillation alone (ablation) | student 2D pose network · MuPoTS-3D 2D keypoints | 2D pose accuracy | Improved |
Closest papers by meaning
- Dual Networks Based 3D Multi-Person Pose Estimation From Monocular Video (TPAMI 2022 · similarity 0.92)
- Direct Multi-view Multi-person 3D Pose Estimation (NeurIPS 2021 · similarity 0.91)
- TesseTrack: End-to-End Learnable Multi-Person Articulated 3D Pose Tracking (CVPR 2021 · similarity 0.90)
- Monocular 3D Pose and Shape Estimation of Multiple People in Natural Scenes - The Importance of Multiple Scene Constraints (CVPR 2018 · similarity 0.90)
- Multi-View Multi-Person 3D Pose Estimation With Plane Sweep Stereo (CVPR 2021 · similarity 0.90)
- SelfPose3d: Self-Supervised Multi-Person Multi-View 3d Pose Estimation (CVPR 2024 · similarity 0.89)
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.
