Research map / Detection and segmentation / Tracking
SQE: a Self Quality Evaluation Metric for Parameters Optimization in Multi-Object Tracking
Yanru Huang, Feiyu Zhu, Zheni Zeng, Xi Qiu, Yuan Shen, Jianan Wu
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 |
|---|---|---|---|
| Tune REID and merging thresholds per video using SQE instead of fixed training-set parameters | MOT tracker · MOT16 training videos 02, 04, 05, 09 (overall) | IDF1: 74.0 -> 75.5 | Improved |
| Tune REID and merging thresholds per video using SQE instead of fixed training-set parameters | MOT tracker · MOT16 training videos 02, 04, 05, 09 (overall) | MOTA: 68.4 -> 69.5 | Improved |
| Tune REID and merging thresholds per video using SQE instead of fixed training-set parameters | MOT tracker · MOT16 training videos 02, 04, 05, 09 (overall) | IDS: 116 -> 138 | Hurt |
| Optimize tracking parameters using SQE rather than fixed baseline parameters | MOT tracker · MOT16 test | IDF1: 66.6 -> 68.3 | Improved |
| Optimize tracking parameters using SQE rather than fixed baseline parameters | MOT tracker · KITTI train | IDF1: 67.4 -> 68.5 | Improved |
| Use feature-distance distributions and a two-class Gaussian mixture to score trajectory quality | SQE · MOT16-02 | IDF1 trend correlation | Improved |
| Apply SQE-based parameter selection to Deep SORT, excluding features from occlusion frames | Deep SORT · MOT16 | IDF1 trend correlation | Improved |
Closest papers by meaning
- HOTA: A Higher Order Metric for Evaluating Multi-object Tracking (IJCV 2020 · similarity 0.87)
- SOTVerse: A User-Defined Task Space of Single Object Tracking (IJCV 2023 · similarity 0.87)
- Beyond Standard Benchmarks: Parameterizing Performance Evaluation in Visual Object Tracking (ICCV 2017 · similarity 0.86)
- MOTCOM: The Multi-Object Tracking Dataset Complexity Metric (ECCV 2022 · similarity 0.86)
- Self-Supervised Multi-Object Tracking with Path Consistency (CVPR 2024 · similarity 0.85)
- Hybrid-SORT: Weak Cues Matter for Online Multi-Object Tracking (AAAI 2024 · similarity 0.84)
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.
