Research map / Detection and segmentation / Tracking
Towards Sequence-Level Training for Visual Tracking
Minji Kim, Seungkwan Lee, Jungseul Ok, Bohyung Han, Minsu Cho
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 sequence-level sampling (SS) (ablation) | SiamRPN++ · LaSOT | AUC (%): 51.0 -> 55.1 | Improved |
| add sequence-level sampling (SS) (ablation) | SiamRPN++ · GOT-10k val | AO (%): 66.4 -> 70.2 | Improved |
| add sequence-level objective (SO) on top of SS (ablation) | SiamRPN++ · LaSOT | AUC (%): 55.1 -> 57.3 | Improved |
| add sequence-level objective (SO) on top of SS (ablation) | SiamRPN++ · GOT-10k val | AO (%): 70.2 -> 73.8 | Improved |
| add sequence-level augmentation (SA) on top of SS+SO (ablation) | SiamRPN++ · LaSOT | AUC (%): 57.3 -> 58.4 | Improved |
| add sequence-level augmentation (SA) on top of SS+SO (ablation) | SiamRPN++ · GOT-10k val | AO (%): 73.8 -> 74.3 | Improved |
| train with SLT | TrDiMP · TrackingNet test | AUC (%): 78.1 (+0.0) | No clear effect |
| train with SLT | SiamRPN++ · LaSOT test | AUC (%): 51.0 -> 58.4 | Improved |
| train with SLT | TransT · LaSOT test | AUC (%): 64.2 -> 66.8 | Improved |
| train with SLT | TrDiMP · GOT-10k test | AO (%): 67.1 -> 67.5 | Improved |
| train sequences of length 24 instead of 1 (ablation) | SLT-SiamRPN++ · GOT-10k val | AO (%): 65.8 -> 73.8 | Improved |
| longer sequence T=32 instead of 24 (ablation) | SLT-SiamRPN++ · GOT-10k val | AO (%): 73.8 -> 73.4 | No clear effect |
Closest papers by meaning
- Visual Tracking via Hierarchical Deep Reinforcement Learning (AAAI 2021 · similarity 0.90)
- Online Decision Based Visual Tracking via Reinforcement Learning (NeurIPS 2020 · similarity 0.90)
- SeqTrack: Sequence to Sequence Learning for Visual Object Tracking (CVPR 2023 · similarity 0.88)
- End-to-end Active Object Tracking via Reinforcement Learning (ICML 2018 · similarity 0.88)
- Learning Policies for Adaptive Tracking With Deep Feature Cascades (ICCV 2017 · similarity 0.88)
- Autoregressive Sequential Pretraining for Visual Tracking (CVPR 2025 · similarity 0.87)
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.
