Research map / Multi-agent systems and games / Cooperative multi-agent RL
Dual-Agent Deep Reinforcement Learning for Deformable Face Tracking
Minghao Guo, Jiwen Lu, Jie Zhou
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 message passing from the tracking agent to the alignment agent (ablation) | DADRL alignment agent · 300-VW Category 3, video #517 | averaged point-to-point error: 3.29 -> 2.75 | Improved |
| add message passing from the tracking agent to the alignment agent (ablation) | DADRL alignment agent · 300-VW Category 3, video #526 | averaged point-to-point error: 3.60 -> 2.92 | Improved |
| add message passing from the tracking agent to the alignment agent (ablation) | DADRL alignment agent · 300-VW Category 3, video #528 | averaged point-to-point error: 3.70 -> 3.01 | Improved |
| add message passing from the tracking agent to the alignment agent (ablation) | DADRL alignment agent · 300-VW Category 3, video #533 | averaged point-to-point error: 3.78 -> 3.68 | Improved |
| add the received message to the alignment agent on occluded or motion-blurred videos (ablation) | DADRL alignment agent · 300-VW Category 3 occlusion/motion-artifact videos | alignment error: about 2% performance improvement | Improved |
| add the message from the alignment agent to the tracking agent (3D pose cue) (ablation) | DADRL tracking agent · 300-VW Category 3 | success rate (bounding box) | Improved |
| add communicated messages instead of zeroing them at test time (DADRL vs DADRL-zero) (ablation) | DADRL · 300-VW Category 3, 49 inner points | RMSE / CED | Improved |
| train with 3D projected landmark annotations (DADRL-3D vs DADRL) | DADRL · 300-VW Category 3, 49 inner points | RMSE / CED | Improved |
Closest papers by meaning
- Visual Tracking via Hierarchical Deep Reinforcement Learning (AAAI 2021 · similarity 0.86)
- Deep Reinforcement Learning with Iterative Shift for Visual Tracking (ECCV 2018 · similarity 0.85)
- Empowering Embodied Visual Tracking with Visual Foundation Models and Offline RL (ECCV 2024 · similarity 0.85)
- Tracking as Online Decision-Making: Learning a Policy From Streaming Videos With Reinforcement Learning (ICCV 2017 · similarity 0.85)
- Attention-Aware Deep Reinforcement Learning for Video Face Recognition (ICCV 2017 · similarity 0.84)
- Zero Shot Generalization of Vision-Based RL Without Data Augmentation (ICML 2025 · similarity 0.83)
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.
