Research map / Detection and segmentation / Tracking
Collaborative Deep Reinforcement Learning for Multi-Object Tracking
Liangliang Ren, Jiwen Lu, Zifeng Wang, Qi Tian, 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 the learned decision network instead of Hungarian algorithm (ablation) | C-DRL · MOT15 train (SubCNN detections) | MOTA: 54.9 -> 67.8 | Improved |
| add the learned decision network instead of Hungarian algorithm (ablation) | C-DRL · MOT15 train (SubCNN detections) | ID switches: 2048 -> 307 | Improved |
| add the prediction network instead of velocity model (ablation) | C-DRL · MOT15 train (SubCNN detections) | MOTA: 63.3 -> 67.8 | Improved |
| add the prediction network instead of velocity model (ablation) | C-DRL · MOT15 train (SubCNN detections) | ID switches: 453 -> 307 | Improved |
| add MDNet features instead of color histogram (ablation) | C-DRL · MOT15 train (SubCNN detections) | MOTA: 65.9 -> 67.8 | Improved |
| raise inter-frame relation threshold from 1 to 8 (ablation) | C-DRL · MOT15 train (SubCNN detections) | MOTA: 63 -> 67 | Improved |
| raise inter-frame threshold beyond 8 (ablation) | C-DRL · MOT15 train (SubCNN detections) | MOTA: 67 -> 67 | No clear effect |
| raise inter-frame threshold (ablation) | C-DRL · MOT15 train (SubCNN detections) | FP versus FN | Mixed |
| set gamma 0.8 and beta 0.4 for reward (ablation) | C-DRL · MOT15 train | normalized reward | Improved |
| C-DRL versus AMIR15 online tracker | C-DRL · MOT15 test (public detections) | MOTA: 37.1 vs 37.6 | Mixed |
| C-DRL versus best online tracker on MOT16 | C-DRL · MOT16 test (public detections) | MOTA: 47.3 vs 47.2 | Improved |
Closest papers by meaning
- Visual Tracking via Hierarchical Deep Reinforcement Learning (AAAI 2021 · similarity 0.90)
- Collaborative Deep Reinforcement Learning for Joint Object Search (CVPR 2017 · similarity 0.89)
- Online Decision Based Visual Tracking via Reinforcement Learning (NeurIPS 2020 · similarity 0.89)
- Action-Decision Networks for Visual Tracking With Deep Reinforcement Learning (CVPR 2017 · similarity 0.88)
- End-to-end Active Object Tracking via Reinforcement Learning (ICML 2018 · similarity 0.88)
- Deep Reinforcement Learning for Robotics: A Survey of Real-World Successes (AAAI 2025 · 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.
