Research map / 3D vision / Human pose estimation
PandaNet: Anchor-Based Single-Shot Multi-Person 3D Pose Estimation
Abdallah Benzine, Florian Chabot, Bertrand Luvison, Quoc Cuong Pham, Catherine Achard
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 |
|---|---|---|---|
| Use anchor-based full-pose regression rather than heatmap-based joint prediction | PandaNet · JTA | 3DPCK: 43.9 -> 83.2 | Improved |
| Use anchor-based full-pose regression | PandaNet · CMU Panoptic Mean | MPJPE (mm): 68.5 -> 42.7 | Improved |
| Use PandaNet instead of XNect | MuPoTS-3D · MuPoTS-3D | 3DPCK: 70.4 -> 72.0 | Improved |
| Use Pose-Aware Anchor Selection instead of no anchor selection (ablation) | PandaNet · JTA | 3DPCK: 80.7 -> 83.2 | Improved |
| Use BB-Aware Anchor Selection instead of no anchor selection (ablation) | PandaNet · JTA | 3DPCK: 80.7 -> 81.9 | Improved |
| Use Pose-Aware rather than BB-Aware Anchor Selection (ablation) | PandaNet · JTA | 3DPCK: 81.9 -> 83.2 | Improved |
| Use Pose-Aware rather than BB-Aware Anchor Selection (ablation) | PandaNet · JTA | AP: 85.1 -> 85.3 | No clear effect |
| Learn task-specific loss weights rather than fixing all weights to 1 (ablation) | PandaNet · JTA | 3DPCK: 15.8 -> 80.8 | Improved |
| Learn anchor weights in addition to task weights (ablation) | PandaNet · JTA | 3DPCK: 80.8 -> 81.7 | Improved |
| Learn joint weights in addition to task and anchor weights (ablation) | PandaNet · JTA | 3DPCK: 81.7 -> 83.2 | Improved |
Later papers that compared with this method
1 later accepted paper reports PandaNet as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Reconstructing Groups of People with Hypergraph Relational Reasoning (ICCV 2023 · better on 1 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- LCR-Net++ — LCR-Net++: Multi-person 2D and 3D Pose Detection in Natural Images (TPAMI 2019) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Dual Networks Based 3D Multi-Person Pose Estimation From Monocular Video (TPAMI 2022 · similarity 0.90)
- PandaPose: 3D Human Pose Lifting from a Single Image via Propagating 2D Pose Prior to 3D Anchor Space (NeurIPS 2025 · similarity 0.90)
- Graph and Temporal Convolutional Networks for 3D Multi-person Pose Estimation in Monocular Videos (AAAI 2021 · similarity 0.89)
- A Simple yet Effective Baseline for 3D Human Pose Estimation (ICCV 2017 · similarity 0.89)
- Distribution-Aware Single-Stage Models for Multi-Person 3D Pose Estimation (CVPR 2022 · similarity 0.88)
- Unsupervised Cross-Modal Alignment for Multi-Person 3D Pose Estimation (ECCV 2020 · similarity 0.88)
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.
