Research map / Detection and segmentation / Tracking
Correlation-Guided Attention for Corner Detection Based Visual Tracking
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 template-guided correlation attention instead of omitting template integration (ablation) | CGACD · OTB2015 + UAV123 | PS: 81.8 -> 86.7 | Helped |
| add template-guided correlation attention instead of omitting template integration (ablation) | CGACD · OTB2015 + UAV123 | AUC: 62.0 -> 66.3 | Helped |
| add both spatial and channel attention rather than channel attention alone (ablation) | CGACD · OTB2015 + UAV123 | PS: 83.8 -> 86.7 | Helped |
| add both spatial and channel attention rather than spatial attention alone (ablation) | CGACD · OTB2015 + UAV123 | PS: 85.4 -> 86.7 | Helped |
| use pixel-wise and channel-wise correlation attention instead of concatenation-guided attention (ablation) | CGACD · OTB2015 + UAV123 | PS: 84.0 -> 86.7 | Helped |
| use pixel-wise and channel-wise correlation attention instead of cross-correlation-guided attention (ablation) | CGACD · OTB2015 + UAV123 | PS: 85.2 -> 86.7 | Helped |
| add RoI augmentation during training (ablation) | CGACD with Conv4 features · OTB2015 + UAV123 | PS: 86.7 -> 87.2 | Helped |
| add RoI augmentation during training (ablation) | CGACD with Conv4 features · OTB2015 + UAV123 | AUC: 66.3 -> 66.8 | Helped |
| use Conv4 features instead of Conv3 features (ablation) | CGACD · OTB2015 + UAV123 | PS: 83.5 -> 86.7 | Helped |
| use Conv4 features instead of Conv5 features (ablation) | CGACD · OTB2015 + UAV123 | PS: 83.8 -> 86.7 | Helped |
| add the corner detection module to the Siamese tracking module (ablation) | CGACD · OTB2015 + UAV123 | PS: +4.0% | Helped |
| replace the first-stage tracker with SiamFC and construct the RoI using the previous target size (ablation) | CGACD · OTB2015 + UAV123 | PS: 80.9% -> 85.5% | Helped |
Closest papers by meaning
- Graph Attention Tracking (CVPR 2021 · similarity 0.87)
- Attentional Correlation Filter Network for Adaptive Visual Tracking (CVPR 2017 · similarity 0.87)
- Deformable Siamese Attention Networks for Visual Object Tracking (CVPR 2020 · similarity 0.86)
- Alpha-Refine: Boosting Tracking Performance by Precise Bounding Box Estimation (CVPR 2021 · similarity 0.86)
- Transformer Tracking (CVPR 2021 · similarity 0.86)
- TAPTRv2: Attention-based Position Update Improves Tracking Any Point (NeurIPS 2024 · similarity 0.84)
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.