Research map / Detection and segmentation / Tracking
Learning Support Correlation Filters for Visual Tracking
Wangmeng Zuo, Xiaohe Wu, Liang Lin, Lei Zhang, Ming–Hsuan Yang
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 HOG and color names features instead of raw pixels (ablation) | MSCF · OTB | mean DP (%): 64.9 -> 80.6 | Improved |
| add HOG and color names features instead of raw pixels (ablation) | KSCF · OTB | mean DP (%): 64.4 -> 85.0 | Improved |
| add color names to HOG features (ablation) | KSCF · OTB | mean AUC (%): 53.2 -> 57.5 | Improved |
| add color names to HOG features (ablation) | MSCF · OTB | mean FPS: 64 -> 54 | Hurt |
| use Gaussian kernel instead of linear kernel (ablation) | KSCF · OTB | mean DP (%): 82.0 -> 85.0 | Improved |
| use Gaussian kernel instead of linear kernel (ablation) | KSCF · OTB | mean FPS: 94 -> 35 | Hurt |
| use polynomial kernel instead of linear kernel (ablation) | KSCF · OTB | mean DP (%): 82.0 -> 84.2 | Improved |
| replace DCF with max-margin MSCF (HOG + CN) | correlation filter · OTB | mean DP (%): 76.2 -> 80.6 | Improved |
| replace KCF with max-margin KSCF (HOG + CN) | kernel correlation filter · OTB | mean DP (%): 75.8 -> 85.0 | Improved |
| add scale adaptation to KSCF (SKSCF) | KSCF · OTB | mean DP (%): 85.0 -> 87.4 | Improved |
| add scale adaptation to KSCF (SKSCF) | KSCF · OTB | mean FPS: 35 -> 8 | Hurt |
| use KSCF instead of MEEM | SVM-based trackers · OTB | mean DP (%): 83.3 -> 85.0 | Improved |
Closest papers by meaning
- Structural Correlation Filter for Robust Visual Tracking (CVPR 2016 · similarity 0.88)
- Visual Tracking via Adaptive Spatially-Regularized Correlation Filters (CVPR 2019 · similarity 0.88)
- Real-Time Visual Tracking: Promoting the Robustness of Correlation Filter Learning (ECCV 2016 · similarity 0.87)
- Learning Spatial-Temporal Regularized Correlation Filters for Visual Tracking (CVPR 2018 · similarity 0.87)
- Correlation Tracking via Joint Discrimination and Reliability Learning (CVPR 2018 · similarity 0.87)
- Discriminative Correlation Filter Tracker with Channel and Spatial Reliability (IJCV 2018 · similarity 0.86)
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.
