Research map / Detection and segmentation / Tracking
Tracking by Animation: Unsupervised Learning of Multi-Object Attentive Trackers
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 |
|---|---|---|---|
| Use adaptive computation time rather than constant computation time (ablation) | TBA · Sprites-MOT | IDF1: 99.0 -> 99.2 | Helped |
| Use adaptive computation time rather than constant computation time (ablation) | TBA · MNIST-MOT | IDF1: 99.2 -> 99.6 | Helped |
| Add occlusion modeling with multiple layers (ablation) | TBAc · Sprites-MOT | IDF1: 93.3 -> 99.0 | Helped |
| Add explicit attention for feature association (ablation) | TBAc · Sprites-MOT | IDF1: 43.2 -> 99.0 | Helped |
| Add explicit attention for feature association (ablation) | TBAc · MNIST-MOT | IDF1: 45.2 -> 99.2 | Helped |
| Enable memory writes so trackers can interact (ablation) | TBAc · Sprites-MOT | IDF1: 0 -> 99.0 | Helped |
| Enable memory writes so trackers can interact (ablation) | TBAc · MNIST-MOT | IDF1: 0 -> 99.2 | Helped |
| Add tracker reprioritization by confidence ranking (ablation) | TBAc · Sprites-MOT | IDF1: 93.0 -> 99.0 | Helped |
| Add tracker reprioritization by confidence ranking (ablation) | TBAc · MNIST-MOT | IDF1: 94.3 -> 99.2 | Helped |
| Use temporal tracking rather than single-image inference | TBA · Sprites-MOT | identity tracking | Helped |
| Use temporal tracking rather than single-image inference | TBA · MNIST-MOT | object disambiguation under overlap | Helped |
Closest papers by meaning
- DiffusionTrack: Diffusion Model for Multi-Object Tracking (AAAI 2024 · similarity 0.90)
- Multiple Object Tracking as ID Prediction (CVPR 2025 · similarity 0.89)
- Focusing on Tracks for Online Multi-Object Tracking (CVPR 2025 · similarity 0.89)
- Video Diffusion Models Excel at Tracking Similar-Looking Objects Without Supervision (NeurIPS 2025 · similarity 0.89)
- Track-On: Transformer-based Online Point Tracking with Memory (ICLR 2025 · similarity 0.88)
- Hybrid-SORT: Weak Cues Matter for Online Multi-Object Tracking (AAAI 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.