Research map / Detection and segmentation / Tracking
Model Uncertainty Guides Visual Object Tracking
Lijun Zhou, Antoine Ledent, Qintao Hu, Ting Liu, Jianlin Zhang, Marius Kloft
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 |
|---|---|---|---|
| Use uncertainty-adjusted S3 representative-frame selection instead of baseline sample selection (ablation) | Dimp50 · UAV123 | AUC: 0.654 -> 0.671 | Improved |
| Add mixed-feature augmentation (ablation) | Dimp50 · UAV123 | AUC: 0.654 -> 0.663 | Improved |
| Replace backbone 3*3 convolutions with deformable convolutions (ablation) | Dimp50 · UAV123 | AUC: 0.654 -> 0.664 | Improved |
| Combine S3, mixed-feature augmentation, and deformable convolutions (ablation) | Dimp50 · UAV123 | AUC: 0.654 -> 0.676 | Improved |
| Replace S3 with random sample selection (ablation) | Dimp50 · UAV123 | AUC: 0.671 -> 0.662 | Hurt |
| Replace S3 with ECO Gaussian Mixture Model selection (ablation) | Dimp50 · UAV123 | AUC: 0.671 -> 0.656 | Hurt |
| Replace uncertainty-weighted mean with standard mean selection (ablation) | Dimp50 · UAV123 | AUC: 0.671 -> 0.659 | Hurt |
| Replace S3 with direct IoU-prediction sample selection (ablation) | Dimp50 · UAV123 | AUC: 0.671 -> 0.655 | Hurt |
| Add S3 sample updating (ablation) | Dimp50 · LaSOT testing set | success rate: 0.569 -> 0.584 | Improved |
| Add S3 sample updating (ablation) | ATOM · LaSOT testing set | success rate: 0.515 -> 0.532 | Improved |
| Add S3 sample updating (ablation) | UpdateNet · LaSOT testing set | success rate: 0.475 -> 0.490 | Improved |
| Add S3 sample updating (ablation) | ECO · LaSOT testing set | success rate: 0.324 -> 0.391 | Improved |
Closest papers by meaning
- UAST: Uncertainty-Aware Siamese Tracking (ICML 2022 · similarity 0.87)
- ATOM: Accurate Tracking by Overlap Maximization (CVPR 2019 · similarity 0.87)
- Learning Occlusion-Robust Vision Transformers for Real-Time UAV Tracking (CVPR 2025 · similarity 0.86)
- LVPTrack: High Performance Domain Adaptive UAV Tracking with Label Aligned Visual Prompt Tuning (AAAI 2025 · similarity 0.86)
- Learning Adaptive and View-Invariant Vision Transformer for Real-Time UAV Tracking (ICML 2024 · similarity 0.85)
- MUST: The First Dataset and Unified Framework for Multispectral UAV Single Object Tracking (CVPR 2025 · 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.
