Research map / Detection and segmentation / Tracking
Learning of Global Objective for Network Flow in Multi-Object Tracking
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Train association costs end-to-end by back-propagating through multi-frame data association instead of using BCE-only training | LPT · MOT17 validation | MOTA: 49.33 -> 51.54 | Helped |
| Train association costs end-to-end by back-propagating through multi-frame data association instead of using BCE-only training | LPT · MOT17 validation | IDF1: 58.91 -> 60.15 | Helped |
| Use L2 rather than L1 loss for end-to-end association training (ablation) | LPT · MOT17 validation | MOTA: 51.45 -> 51.54 | Helped |
| Use L2 rather than L1 loss for end-to-end association training (ablation) | LPT · MOT17 validation | IDF1: 60.15 -> 60.15 | No clear effect |
| Add second-round MCF for long-term occlusion handling (ablation) | LPT · MOT16 validation | IDF1: 43.5 -> 55.1 | Helped |
| Add second-round MCF for long-term occlusion handling (ablation) | LPT · MOT16 validation | MOTA: 38.9 -> 42.9 | Helped |
| Add second-round MCF for long-term occlusion handling (ablation) | LPT · MOT16 validation | IDS: 134 -> 73 | Helped |
| Replace BCE affinity training with global association-based end-to-end training | LPT · MOT16 validation | MOTA: 31.10 -> 44.27 | Helped |
| Use L2 loss rather than BCE for affinity and association learning (ablation) | Proposed end-to-end affinity learner · MOT16 validation | MSE Edge: 0.017 -> 0.006 | Helped |
| Use L2 loss rather than BCE for affinity and association learning (ablation) | Proposed end-to-end affinity learner · MOT16 validation | AUC: 0.996 -> 0.997 | Helped |
Closest papers by meaning
- Sampling-Resilient Multi-Object Tracking (AAAI 2024 · similarity 0.84)
- DiffusionTrack: Diffusion Model for Multi-Object Tracking (AAAI 2024 · similarity 0.84)
- Multi-Scene Generalized Trajectory Global Graph Solver with Composite Nodes for Multiple Object Tracking (AAAI 2024 · similarity 0.83)
- Temporal Coherent Object Flow for Multi-Object Tracking (AAAI 2025 · similarity 0.83)
- Focusing on Tracks for Online Multi-Object Tracking (CVPR 2025 · similarity 0.83)
- Multiple Object Tracking as ID Prediction (CVPR 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.