Research map / Learning theory and algorithms / Kernels and regression theory
Fast-deepKCF Without Boundary Effect
Linyu Zheng, Ming Tang, Yingying Chen, Jinqiao Wang, Hanqing Lu
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 |
|---|---|---|---|
| avoid boundary effect with real dense samples and deep features (fdKCF* vs BACF) | real-time CF tracker · OTB-2013 | DP: 0.858 -> 0.908 | Improved |
| avoid boundary effect with real dense samples and deep features (fdKCF* vs ECO-HC) | real-time CF tracker · OTB-2015 | DP: 0.845 -> 0.891 | Improved |
| replace cyclic samples (HCF) with real dense samples | correlation filter with deep features · OTB-2013 and OTB-2015 | localization accuracy: about 10 percent point | Improved |
| spatial-domain fCKM instead of FFT-based kernel matrix (fdKCF* vs HCF speed) | deep CF tracker · OTB-2015 | fps: 11 -> 24 | Improved |
| use deep features with spatial-domain fCKM instead of C-COT | deep CF tracker · OTB-2015 | fps: 0.3 -> 24 | Improved |
| add hand-crafted HOG and Color-Names features to deep features (ECO vs fdKCF*) | deep CF tracker · OTB-2015 very hard sequences | DP: 0.789 -> 0.717 | Hurt |
| add hand-crafted HOG and Color-Names features, GMM clustering and dimensionality reduction (ECO vs fdKCF*) | deep CF tracker · OTB-2015 | DP: 0.891 -> 0.897 | No clear effect |
| use deep features only versus ECO on OTB-2013 success AUC | deep CF tracker · OTB-2013 | AUC: 0.705 -> 0.711 | No clear effect |
Closest papers by meaning
- High-Speed Tracking With Multi-Kernel Correlation Filters (CVPR 2018 · similarity 0.90)
- Context-Aware Correlation Filter Tracking (CVPR 2017 · similarity 0.87)
- Learning Background-Aware Correlation Filters for Visual Tracking (ICCV 2017 · similarity 0.87)
- Visual Tracking via Adaptive Spatially-Regularized Correlation Filters (CVPR 2019 · similarity 0.86)
- Discriminative Correlation Filter Tracker with Channel and Spatial Reliability (IJCV 2018 · similarity 0.86)
- Discriminative Correlation Filter With Channel and Spatial Reliability (CVPR 2017 · 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.
