Research map / Detection and segmentation / Tracking
Attention to Trajectory: Trajectory-Aware Open-Vocabulary Tracking
Yunhao Li, Yifan Jiao, Dan Meng, Heng Fan, Libo Zhang
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 TCR strategy (ablation) | TRACT baseline · OV-TAO val (YOLO-World) | ClsA: 16.9 -> 17.3 | Improved |
| add TFA on top of TCR (ablation) | TRACT · OV-TAO val (YOLO-World) | ClsA: 17.3 -> 19.9 | Improved |
| add TSE on top of TCR+TFA (ablation) | TRACT · OV-TAO val (YOLO-World) | ClsA: 19.9 -> 20.3 | Improved |
| add TCR on top of TFA+TSE (ablation) | TRACT · OV-TAO val (YOLO-World) | TETA: 38.4 -> 38.6 | Improved |
| add TSE to base-class classification (ablation) | TRACT · OV-TAO val base classes | ClsA: 21.6 -> 20.2 | Hurt |
| add TSE to novel-class classification (ablation) | TRACT · OV-TAO val novel classes | ClsA: 10.9 -> 13.3 | Improved |
| increase n_clip from 1 to 5 (ablation) | TraCLIP · OV-TAO val | ClsA: 19.09 -> 20.30 | Improved |
| increase n_clip from 5 to 15 (ablation) | TraCLIP · OV-TAO val | ClsA: 20.30 -> 20.13 | No clear effect |
| increase n_bank from 5 to 15 (ablation) | TCR · OV-TAO val | ClsA: 16.52 -> 17.27 | Mixed |
| increase alpha from 0.1 to 0.25 (ablation) | TCR · OV-TAO val | TETA: 37.41 -> 37.62 | Improved |
| increase alpha from 0.25 to 0.4 (ablation) | TCR · OV-TAO val | TETA: 37.62 -> 37.17 | Hurt |
| use TRACT instead of MASA with YOLO-World | TRACT · OV-TAO val novel | ClsA: 4.4 -> 5.3 | Improved |
Closest papers by meaning
- OVTR: End-to-End Open-Vocabulary Multiple Object Tracking with Transformer (ICLR 2025 · similarity 0.89)
- OVT-B: A New Large-Scale Benchmark for Open-Vocabulary Multi-Object Tracking (NeurIPS 2024 · similarity 0.88)
- DeconfuseTrack: Dealing with Confusion for Multi-Object Tracking (CVPR 2024 · similarity 0.85)
- Focusing on Tracks for Online Multi-Object Tracking (CVPR 2025 · similarity 0.85)
- Multiple Object Tracking as ID Prediction (CVPR 2025 · similarity 0.85)
- Towards Generalizable Multi-Object Tracking (CVPR 2024 · 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.
