Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Meta Agent Teaming Active Learning for Pose Estimation
Jia Gong, Zhipeng Fan, Qiuhong Ke, Hossein Rahmani, Jun Liu
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 MATAL to select images for annotation | pose estimation active learning · ICVL | average joint error | Improved |
| use MATAL to select images for annotation | pose estimation active learning · MPII | labeling effort at similar PCKh@0.5: 4K images; 6K images; around 2K images | Improved |
| use MATAL to select images for annotation | pose estimation active learning · NYU | average joint error: 5K images; 23.5 mm; around 10K labeled images | Improved |
| add global topological feature KFP0 to the state (ablation) | MATAL · NYU | MSE (mm), 8000 labeled samples: 23.01 -> 22.53 | Improved |
| add local topological features to the state (ablation) | MATAL · NYU | MSE (mm), 8000 labeled samples: 23.23 -> 22.53 | Improved |
| add appearance feature KFA to the state (ablation) | MATAL · NYU | MSE (mm), 8000 labeled samples: 23.55 -> 22.53 | Improved |
| add budget consumption ratio b to the state (ablation) | MATAL · NYU | MSE (mm), 8000 labeled samples: 23.85 -> 22.53 | Improved |
| add topological novelty scores to the action (ablation) | MATAL · NYU | MSE (mm), 8000 labeled samples: 25.13 -> 22.53 | Improved |
| add topological representativeness histograms to the action (ablation) | MATAL · NYU | MSE (mm), 8000 labeled samples: 24.12 -> 22.53 | Improved |
| add appearance feature fA to the action (ablation) | MATAL · NYU | MSE (mm), 8000 labeled samples: 23.69 -> 22.53 | Improved |
| add meta-optimization for agent-team retraining (ablation) | MATAL · NYU | time cost: 4.5 -> 2 | Improved |
| add meta-optimization for agent-team retraining (ablation) | MATAL · NYU | MSE (mm): 23.68 -> 23.74 | Hurt |
Closest papers by meaning
- Active Learning for Human Pose Estimation (ICCV 2017 · similarity 0.88)
- Deep Reinforcement Learning for Active Human Pose Estimation (AAAI 2020 · similarity 0.87)
- MAT-Agent: Adaptive Multi-Agent Training Optimization (NeurIPS 2025 · similarity 0.84)
- Multi-agent Long-term 3D Human Pose Forecasting via Interaction-aware Trajectory Conditioning (CVPR 2024 · similarity 0.83)
- Fast Adaptation for Human Pose Estimation via Meta-Optimization (CVPR 2024 · similarity 0.83)
- GRPose: Learning Graph Relations for Human Image Generation with Pose Priors (AAAI 2025 · similarity 0.82)
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.
