Research map / Detection and segmentation / Tracking
Discriminative Scale Space Tracking
Martin Danelljan, Gustav Hager, Fahad Shahbaz Khan, Michael Felsberg
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 a separate scale filter (DSST) to the translation DCF (ablation) | translation DCF tracker · OTB-50 | Mean OP (%): 57.7 -> 67.7 | Improved |
| add the scale filter to the translation DCF (ablation) | translation DCF tracker · OTB-50 | AUC (%): 49.5 -> 6.6% gain | Improved |
| add the scale filter to the translation DCF (ablation) | translation DCF tracker · VOT 2014 | Final rank: 3.31 -> 2.80 | Improved |
| add the scale filter to the translation DCF (ablation) | translation DCF tracker · VOT 2014 | Average failures: 1.48 -> 1.16 | Improved |
| use separate 1D scale filter instead of joint scale space filter | DSST · OTB-50 | Mean FPS: 1.46 -> 25.4 | Improved |
| use separate scale filter instead of joint scale space filter | DSST · OTB-50 | Mean OP (%): 63.2 -> 67.7 | Improved |
| use separate scale filter instead of multi-resolution translation filter | DSST · OTB-50 | Mean OP (%): 65.2 -> 67.7 | Improved |
| add iterative refinement to joint scale space filter | Joint DCF · OTB-50 | Mean OP (%): 63.2 -> 64.1 | Improved |
| add PCA dimensionality reduction, faster features and larger search space (fDSST) | DSST · OTB-50 | Mean OP (%): 67.7 -> 74.3 | Improved |
| add PCA dimensionality reduction, faster features and larger search space (fDSST) | DSST · OTB-50 | Mean FPS: 25.4 -> 54.3 | Improved |
| reduce translation filter subspace dimensions with PCA (ablation) | fDSST · OTB-50 | AUC: 32 -> 18 | No clear effect |
| use fDSST with explicit scale filter instead of multi-feature SAMF | SAMF · OTB-50 | Mean OP (%): 69.7 -> 74.3 | Improved |
Closest papers by meaning
- Adaptive Decontamination of the Training Set: A Unified Formulation for Discriminative Visual Tracking (CVPR 2016 · similarity 0.87)
- Adaptive Correlation Filters with Long-Term and Short-Term Memory for Object Tracking (IJCV 2018 · similarity 0.87)
- Learning Discriminative Model Prediction for Tracking (ICCV 2019 · similarity 0.86)
- Visual Tracking via Adaptive Spatially-Regularized Correlation Filters (CVPR 2019 · similarity 0.86)
- ATOM: Accurate Tracking by Overlap Maximization (CVPR 2019 · similarity 0.85)
- OmniTracker: Unifying Visual Object Tracking by Tracking-With-Detection (TPAMI 2025 · 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.
