Research map / 3D vision / Human pose estimation
On the Benefits of 3D Pose and Tracking for Human Action Recognition
Jathushan Rajasegaran, Georgios Pavlakos, Angjoo Kanazawa, Christoph Feichtenhofer, Jitendra Malik
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 person tracking to MViT appearance features (ablation) | MViT action recognition on AVA · AVA | mAP: 40.2 -> 41.4 (+1.2) | Improved |
| Add 3D pose features to tracked MViT features (ablation) | MViT + Tracking · AVA | mAP: 41.4 -> 42.3 (+0.9) | Improved |
| Add 3D pose and tracking to appearance features | MViT on AVA v2.2 · AVA v2.2 | mAP: +2.8 mAP | Improved |
| Add LART track modeling to Hiera appearance features | Hiera on AVA v2.2 · AVA v2.2 | mAP: +2.5 mAP | Improved |
| Model other people's 3D tracks alongside the person-of-interest (ablation) | LART-pose · AVA | Person Interactions mAP: +2.1 mAP | Improved |
| Model five people's 3D tracks instead of only the target person's track (ablation) | LART-pose · AVA | mAP: 22.9 -> 24.1; 2.0 mAP | Improved |
| Increase the number of people in the pose-model context (ablation) | LART-pose · AVA | mAP | Improved |
| Use explicit 3D pose and tracking instead of appearance-only MViT features | LART-pose versus JMRN · AVA | mAP: +10.0 mAP | Improved |
Closest papers by meaning
- PA3D: Pose-Action 3D Machine for Video Recognition (CVPR 2019 · similarity 0.86)
- LA-Pose: Latent Action Pretraining Meets Pose Estimation (CVPR 2026 · similarity 0.86)
- PoseTrack: A Benchmark for Human Pose Estimation and Tracking (CVPR 2018 · similarity 0.86)
- Deep Reinforcement Learning for Active Human Pose Estimation (AAAI 2020 · similarity 0.86)
- First-Person Hand Action Benchmark With RGB-D Videos and 3D Hand Pose Annotations (CVPR 2018 · similarity 0.86)
- FreeMan: Towards Benchmarking 3D Human Pose Estimation under Real-World Conditions (CVPR 2024 · similarity 0.85)
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.
