Research map / Detection and segmentation / Tracking
PVT++: A Simple End-to-End Latency-Aware Visual Tracking Framework
Bowen Li, Ziyuan Huang, Junjie Ye, Yiming Li, Sebastian Scherer, Hang Zhao, Changhong Fu
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 |
|---|---|---|---|
| predict relative motion factor instead of raw motion value (ablation) | PVT++ motion predictor · DTB70 | AUC@La0: 0.300 -> 0.385 | Improved |
| add auxiliary supervision branch (ablation) | PVT++ P_MV · DTB70 | AUC@La0: 0.323 -> 0.399 | Improved |
| train tracker and predictor jointly instead of fixing the tracker (ablation) | PVT++ P_MV · DTB70 | AUC@La0: 0.294 -> 0.399 | Improved |
| add visual cues to the motion predictor (joint model) (ablation) | PVT++ · DTB70 | AUC@La0: 0.385 -> 0.399 | Improved |
| increase predictor channel width to 128 (ablation) | PVT++ motion encoder · DTB70 | mAUC: 0.399 -> 0.373 | Hurt |
| replace Kalman filter with PVT++ learned predictor | SiamRPN++_M · four UAV datasets averaged | AUC@La0: 0.462 -> 0.505 | Improved |
| add PVT++ latency-aware predictor to offline-accurate trackers | SiamRPN++_R · four UAV datasets averaged | AUC@La0: 0.287 -> 0.388 | Improved |
| motion-only predictor versus visual-only predictor on small and occluded objects (ablation) | PVT++ · UAVDT attributes | AUC@La0: SO 0.549 (motion) vs 0.525 (visual) | Mixed |
Closest papers by meaning
- Visual Object Tracking for Unmanned Aerial Vehicles: A Benchmark and New Motion Models (AAAI 2017 · similarity 0.90)
- Learning Adaptive and View-Invariant Vision Transformer for Real-Time UAV Tracking (ICML 2024 · similarity 0.89)
- Similarity-Guided Layer-Adaptive Vision Transformer for UAV Tracking (CVPR 2025 · similarity 0.88)
- Learning Occlusion-Robust Vision Transformers for Real-Time UAV Tracking (CVPR 2025 · similarity 0.87)
- LVPTrack: High Performance Domain Adaptive UAV Tracking with Label Aligned Visual Prompt Tuning (AAAI 2025 · similarity 0.86)
- Open-World Drone Active Tracking with Goal-Centered Rewards (NeurIPS 2025 · 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.
