Research map / Detection and segmentation / Tracking
OmniTracker: Unifying Visual Object Tracking by Tracking-With-Detection
Junke Wang, Zuxuan Wu, Dongdong Chen, Chong Luo, Xiyang Dai, Lu Yuan, Yu–Gang Jiang
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 Reference-guided Feature Enhancement (RFE) (ablation) | OmniTracker-T · TrackingNet | Normalized precision: 80.9 -> 82.5 | Improved |
| add RFE (ablation) | OmniTracker-T · DAVIS17 | J&F: 64.7 -> 66.2 | Improved |
| add RFE (ablation) | OmniTracker-T · MOT17 val | MOTA: 72.4 -> 73.2 | Improved |
| add RFE (ablation) | OmniTracker-T · YTVIS19 | mAP: 58.5 -> 58.8 | Improved |
| add RFE, evaluated on DAVIS17 videos outside COCO categories (ablation) | OmniTracker-T · DAVIS17 (non-COCO categories) | J&F: 59.7 -> 61.9 | Improved |
| add ReID loss (ablation) | OmniTracker-T · TrackingNet | Normalized precision: 50.0 -> 82.5 | Improved |
| add ReID loss (ablation) | OmniTracker-T · YTVIS19 | mAP: 35.8 -> 58.8 | Improved |
| add multi-frame memory bank (ablation) | OmniTracker-T · DAVIS17 | J&F: 62.8 -> 66.2 | Improved |
| add Kalman filter at inference (ablation) | OmniTracker-T · MOT17 val | MOTA: 59.7 -> 73.2 | Improved |
| joint training across tasks instead of separate training (ablation) | OmniTracker-T · DAVIS17 | J&F: 65.7 -> 66.2 | Improved |
| joint training instead of separate training (ablation) | OmniTracker-T · TrackingNet | Normalized precision: 82.0 -> 82.5 | Improved |
| ConvNeXt-Large backbone instead of Swin-Large (ablation) | OmniTracker-L · MOT17 | MOTA: 79.1 -> 79.3 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- SMILEtrack — SMILEtrack: SiMIlarity LEarning for Occlusion-Aware Multiple Object Tracking (AAAI 2024) · better on 0 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- VOVTrack: Exploring the Potentiality in Raw Videos for Open-Vocabulary Multi-Object Tracking (ICCV 2025 · similarity 0.92)
- OVTrack: Open-Vocabulary Multiple Object Tracking (CVPR 2023 · similarity 0.91)
- VastTrack: Vast Category Visual Object Tracking (NeurIPS 2024 · similarity 0.91)
- OVT-B: A New Large-Scale Benchmark for Open-Vocabulary Multi-Object Tracking (NeurIPS 2024 · similarity 0.90)
- DiffusionTrack: Diffusion Model for Multi-Object Tracking (AAAI 2024 · similarity 0.88)
- Multiple Object Tracking as ID Prediction (CVPR 2025 · similarity 0.88)
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.
