Research map / Detection and segmentation / Open-set and few-shot detection
Robust Visual Tracking by Segmentation
Matthieu Paul, Martin Danelljan, Christoph Mayer, 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 instance-branch conditioning of the segmentation decoder (ablation) | RTS · LaSOT | AUC: 65.3 -> 69.7 | Improved |
| add instance-branch conditioning (ablation) | RTS · UAV123 | AUC: 65.2 -> 67.6 | Improved |
| add instance-branch conditioning (ablation) | RTS · NFS | AUC: 65.8 -> 65.4 | No clear effect |
| adapt LWL to tracking (RTS without instance branch) | LWL · LaSOT | AUC: 59.7 -> 65.3 | Improved |
| add instance-branch fallback when segmentation cannot localize target (ablation) | RTS · UAV123 | AUC: 66.3 -> 67.6 | Improved |
| add instance-branch fallback (ablation) | RTS · LaSOT | AUC: 69.3 -> 69.7 | Improved |
| lower confidence threshold t_sc from 0.30 to 0.20 (ablation) | RTS · LaSOT | AUC: 69.7 -> 68.6 | Hurt |
| raise confidence threshold t_sc from 0.30 to 0.40 (ablation) | RTS · NFS | AUC: 65.4 -> 63.3 | Hurt |
| add auxiliary instance localization loss (eta=10) (ablation) | RTS · LaSOT | AUC: 67.7 -> 69.7 | Improved |
| increase eta from 0.4 to 10 (ablation) | RTS · LaSOT | AUC: 69.8 -> 69.7 | No clear effect |
| increase eta from 0.4 to 10 (ablation) | RTS · GOT-10k | AO: 84.0 -> 85.2 | Improved |
| fine-tune on YouTube-VOS with longer init and no scale change limit | RTS · YouTube-VOS 2019 | G: 79.7 -> 80.3 | 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.
- Siam-RCNN — 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
- Generating Masks From Boxes by Mining Spatio-Temporal Consistencies in Videos (ICCV 2021 · similarity 0.89)
- A Discriminative Single-Shot Segmentation Network for Visual Object Tracking (TPAMI 2021 · similarity 0.89)
- Track and Segment: An Iterative Unsupervised Approach for Video Object Proposals (CVPR 2016 · similarity 0.89)
- Integrating Boxes and Masks: A Multi-Object Framework for Unified Visual Tracking and Segmentation (ICCV 2023 · similarity 0.88)
- When Trackers Fail: VLM-Guided Verification and Recovery for Robust Video Object Segmentation (NeurIPS 2026 · similarity 0.88)
- TrackIME: Enhanced Video Point Tracking via Instance Motion Estimation (NeurIPS 2024 · 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.
