Research map / Detection and segmentation / Tracking
Learning Target Candidate Association To Keep Track of What Not To Track
Christoph Mayer, Martin Danelljan, Danda Pani Paudel, Luc Van Gool
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 target candidate association network (ablation) | SuperDiMP with memory confidence and search area adaptation · LaSOT | AUC (%): 65.8 -> 67.1 | Improved |
| add target candidate association network (ablation) | SuperDiMP with memory confidence and search area adaptation · NFS | AUC (%): 65.2 -> 66.4 | Improved |
| add target candidate association network (ablation) | SuperDiMP with memory confidence and search area adaptation · UAV123 | AUC (%): 69.1 -> 69.7 | Improved |
| add memory sample confidence (ablation) | SuperDiMP · LaSOT | AUC (%): 63.5 -> 65.0 | Improved |
| add memory sample confidence (ablation) | SuperDiMP · NFS / UAV123 | AUC (%): 64.4 -> 64.7 (NFS); 68.2 -> 68.0 (UAV123) | No clear effect |
| add search area adaptation (larger area and rescaling) (ablation) | SuperDiMP with memory confidence · LaSOT | AUC (%): 65.0 -> 65.8 | Improved |
| add larger search area only (ablation) | SuperDiMP with memory confidence · LaSOT | AUC (%): 65.0 -> 65.5 | Improved |
| add search area rescaling only (ablation) | SuperDiMP with memory confidence · LaSOT | AUC (%): 65.0 -> 65.8 | Improved |
| add partially supervised loss only for association training (ablation) | TCA · LaSOT | AUC (%): 65.8 -> 66.0 | Improved |
| add self-supervised loss only for association training (ablation) | TCA · LaSOT | AUC (%): 65.8 -> 66.9 | Improved |
| add data-mining when training with combined loss (ablation) | TCA · LaSOT | AUC (%): 66.8 -> 67.1 | Improved |
| use sample confidence for replacement, then for online learning, then threshold 0.5 (ablation) | memory management · LaSOT | AUC (%): 63.5 -> 64.1 -> 64.6 -> 65.0 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- SiamR-CNN — Siam R-CNN: Visual Tracking by Re-Detection (CVPR 2020) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Know Your Surroundings: Exploiting Scene Information for Object Tracking (ECCV 2020 · similarity 0.88)
- Distractor-Aware Fast Tracking via Dynamic Convolutions and MOT Philosophy (CVPR 2021 · similarity 0.87)
- A Distractor-Aware Memory for Visual Object Tracking with SAM2 (CVPR 2025 · similarity 0.87)
- Tracking the Untrackable: Learning to Track Multiple Cues With Long-Term Dependencies (ICCV 2017 · similarity 0.87)
- ATOM: Accurate Tracking by Overlap Maximization (CVPR 2019 · similarity 0.87)
- Video Diffusion Models Excel at Tracking Similar-Looking Objects Without Supervision (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.
