Research map / Learning theory and algorithms / Kernels and regression theory
Object Tracking via Dual Linear Structured SVM and Explicit Feature Map
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 |
|---|---|---|---|
| Add the unary explicit feature map to approximate the intersection kernel (ablation) | DLSSVM · TB50 OPE | success (AUC): 0.557 -> 0.589 | Helped |
| Add the unary explicit feature map to approximate the intersection kernel (ablation) | DLSSVM · TB50 OPE | precision (20 pixels): 0.794 -> 0.829 | Helped |
| Add multi-scale estimation (ablation) | DLSSVM · TB50 OPE | success (AUC): 0.589 -> 0.608 | Helped |
| Add multi-scale estimation (ablation) | DLSSVM · TB50 OPE | precision (20 pixels): 0.829 -> 0.861 | Helped |
| Vary support-vector budget from 100 to 50 or 500 (ablation) | DLSSVM · TB50 OPE | success (AUC): 0.589; 0.587; 0.588 | No clear effect |
| Use a closed-form dual-derived update step size instead of a manually selected sub-gradient step size (ablation) | Linear SSVM tracker · TB50 OPE | success (AUC): 0.443 -> 0.589 | Helped |
| Use DLSSVM optimization instead of linear Struck optimization | Linear SSVM tracker · TB50 OPE | success (AUC): 0.556 -> 0.589 | Helped |
| Use multi-scale estimation (ablation) | DLSSVM · TB100 OPE | success (AUC): 0.541 -> 0.563 | Helped |
| Use multi-scale estimation (ablation) | DLSSVM · TB100 OPE | precision: 0.767 -> 0.805 | Helped |
| Use DLSSVM rather than HCFT | Object tracking · TB100 OPE | precision: 0.837 -> 0.767 | Hurt |
Closest papers by meaning
- Large Margin Object Tracking With Circulant Feature Maps (CVPR 2017 · similarity 0.87)
- Release the Power of Online-Training for Robust Visual Tracking (AAAI 2020 · similarity 0.83)
- Fast Algorithms for Linear and Kernel SVM+ (CVPR 2016 · similarity 0.83)
- SPM-Tracker: Series-Parallel Matching for Real-Time Visual Object Tracking (CVPR 2019 · similarity 0.83)
- VastTrack: Vast Category Visual Object Tracking (NeurIPS 2024 · similarity 0.83)
- Event Stream-based Visual Object Tracking: A High-Resolution Benchmark Dataset and A Novel Baseline (CVPR 2024 · similarity 0.81)
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.