Research map / Architectures and efficiency / Attention and transformers
Transformer 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 |
|---|---|---|---|
| replace correlation-based fusion with ECA self-attention and CFA cross-attention (ablation) | TransT · LaSOT, TrackingNet, GOT-10k | AUC (LaSOT/TrackingNet); AO (GOT-10k): 57.7 -> 64.9; 77.5 -> 81.4; 62.8 -> 72.3 | Helped |
| add post-processing window penalty (ablation) | TransT · LaSOT, TrackingNet, GOT-10k | AUC (LaSOT/TrackingNet); AO (GOT-10k): 62.9 -> 64.9; 81.1 -> 81.4; 71.5 -> 72.3 | Helped |
| use the proposed fusion architecture instead of the original Transformer structure (ablation) | TransT · LaSOT, TrackingNet, GOT-10k | AUC (LaSOT/TrackingNet); AO (GOT-10k): 62.3 -> 64.9; 81.3 -> 81.4; 70.3 -> 72.3 | Helped |
| add ECA self-attention while retaining correlation instead of CFA (ablation) | TransT with correlation · LaSOT, TrackingNet, GOT-10k | AUC (LaSOT/TrackingNet); AO (GOT-10k): 47.7 -> 57.7; 68.8 -> 77.5; 50.9 -> 62.8 | Helped |
| add CFA cross-attention while retaining correlation, without ECA (ablation) | TransT without ECA · LaSOT, TrackingNet, GOT-10k | AUC (LaSOT/TrackingNet); AO (GOT-10k): 47.7 -> 62.9; 68.8 -> 81.1; 50.9 -> 70.6 | Helped |
| add ECA self-attention to CFA-only fusion (ablation) | TransT with CFA · LaSOT, TrackingNet, GOT-10k | AUC (LaSOT/TrackingNet); AO (GOT-10k): 62.9 -> 64.9; 81.1 -> 81.4; 70.6 -> 72.3 | Helped |
| train only on GOT-10k rather than the combined datasets | TransT-GOT · GOT-10k test | AO: 72.3 -> 67.1 | Hurt |
Closest papers by meaning
- Compact Transformer Tracker with Correlative Masked Modeling (AAAI 2023 · similarity 0.89)
- Graph Attention Tracking (CVPR 2021 · similarity 0.89)
- Cross-modulated Attention Transformer for RGBT Tracking (AAAI 2025 · similarity 0.87)
- Target Focused Shallow Transformer Framework for Efficient Visual Tracking (AAAI 2024 · similarity 0.86)
- Sequential Fusion Based Multi-Granularity Consistency for Space-Time Transformer Tracking (AAAI 2024 · similarity 0.86)
- Learning Random Noise Salient Feature Fusion Siamese Network for Low-Resolution Object Tracking (Student Abstract) (AAAI 2024 · 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.