Research map / Label-efficient and robust learning / Recognition and long-tail
Efficient Diverse Ensemble for Discriminative Co-Tracking
Kourosh Meshgi, Shigeyuki Oba, Shin Ishii
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 |
|---|---|---|---|
| combine uncertain-sample updates with artificially generated diversity data (ablation) | DEDT ensemble · OTB50 | success AUC | Improved |
| use artificial diversity data alone instead of combining it with tracker samples (ablation) | DEDT ensemble · OTB50 | success AUC | Hurt |
| replace artificial diversity samples with nearest real image patches (ablation) | DEDT · OTB50 | tracking performance | No clear effect |
| query disputed labels from the auxiliary classifier rather than isolate it and fuse outputs afterward (ablation) | DEDT · OTB50 | tracking performance | Improved |
| keep the auxiliary classifier fixed after training on the first frame (ablation) | DEDT · OTB50 | tracking performance | Hurt |
| update the auxiliary classifier every frame, removing its long-memory behavior (ablation) | DEDT · OTB50 | tracking performance | Hurt |
| increase the query threshold so the auxiliary classifier labels more, including slightly uncertain samples (ablation) | DEDT · tracking performance | tracking performance | Hurt |
| use DEDT instead of CCOT | visual tracking · OTB50 | Avg. Succ: 0.75 -> 0.74 | Hurt |
| use DEDT instead of CCOT | visual tracking · OTB100 | Avg. Succ: 0.74 -> 0.69 | Hurt |
| use DEDT instead of SRDCF | visual tracking · VOT2015 | Accuracy: 0.56 -> 0.58 | Improved |
Closest papers by meaning
- DOVTrack: Data-Efficient Open-Vocabulary Tracking (NeurIPS 2025 · similarity 0.84)
- A Unified Theory of Diversity in Ensemble Learning (JMLR 2023 · similarity 0.84)
- ECO: Efficient Convolution Operators for Tracking (CVPR 2017 · similarity 0.84)
- Learning Feature Embeddings for Discriminant Model based Tracking (ECCV 2020 · similarity 0.84)
- DeconfuseTrack: Dealing with Confusion for Multi-Object Tracking (CVPR 2024 · similarity 0.83)
- Multiple Object Tracking as ID Prediction (CVPR 2025 · similarity 0.82)
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.
