Research map / Detection and segmentation / Tracking
DIVOTrack: A Novel Dataset and Baseline Method for Cross-View Multi-Object Tracking in DIVerse Open Scenes
Shengyu Hao, Peiyuan Liu, Yibing Zhan, Kaixun Jin, Zuozhu Liu, Mingli Song, Jenq–Neng Hwang, Gaoang Wang
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 decoupled multi-head embedding (separate single-view and cross-view Re-ID heads) (ablation) | CrossMOT · DIVOTrack easy test set | CVMA: 70.7 -> 72.4 | Improved |
| add decoupled multi-head embedding (ablation) | CrossMOT · DIVOTrack easy test set | CVIDF1: 70.9 -> 71.1 | Improved |
| add conflict-free loss (ablation) | CrossMOT · DIVOTrack easy test set | CVMA: 70.1 -> 72.4 | Improved |
| add conflict-free loss (ablation) | CrossMOT · DIVOTrack easy test set | CVIDF1: 69.8 -> 71.1 | Improved |
| add decoupled multi-head embedding (ablation) | CrossMOT · MvMHAT | CVMA: 92.2 -> 92.3 | Improved |
| add conflict-free loss (ablation) | CrossMOT · MvMHAT | CVIDF1: 87.0 -> 87.4 | Improved |
| add conflict-free loss (ablation) | CrossMOT · CAMPUS | CVMA: 64.4 -> 65.6 | Improved |
| add decoupled multi-head embedding (ablation) | CrossMOT · CAMPUS | CVIDF1: 61.9 -> 61.2 | Hurt |
| vary single-view threshold delta_s from 0.1 to 0.4 with delta_c 0.3 or 0.5 (ablation) | CrossMOT · DIVOTrack easy test set | CVIDF1 | No clear effect |
| use joint detection and embedding (FairMOT) instead of separate detector tracker | FairMOT vs Deepsort/CenterTrack/Tracktor/TraDes · DIVOTrack easy test set | HOTA: 65.3 vs 48.4 to 58.9 | Improved |
| move from easy to hard test set | single-view trackers · DIVOTrack hard test set | HOTA: 65.3 -> 56.5 | Hurt |
| CrossMOT vs best baseline CT | CrossMOT · DIVOTrack easy test set | CVMA: 64.9 -> 72.4 | Improved |
Closest papers by meaning
- Cross-View Referring Multi-Object Tracking (AAAI 2025 · similarity 0.89)
- Multiple Human Association and Tracking from Egocentric and Complementary Top Views (TPAMI 2021 · similarity 0.88)
- Multi-View 3D Point Tracking (ICCV 2025 · similarity 0.88)
- Repurposing Video Diffusion Transformers for Cross-View Temporal Object Correspondence (NeurIPS 2026 · similarity 0.88)
- DiffusionTrack: Diffusion Model for Multi-Object Tracking (AAAI 2024 · similarity 0.87)
- VastTrack: Vast Category Visual Object Tracking (NeurIPS 2024 · similarity 0.87)
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.
