Research map / Detection and segmentation / Tracking
MOTCOM: The Multi-Object Tracking Dataset Complexity Metric
Malte Pedersen, Joakim Bruslund Haurum, Patrick Dendorfer, Thomas B. Moeslund
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 |
|---|---|---|---|
| use MOTCOM instead of density to rank sequences | complexity ranking · MOT17 test | normalized Footrule Distance: 0.50 -> 0.25 | Improved |
| use MOTCOM instead of tracks to rank sequences | complexity ranking · MOT17 test | normalized Footrule Distance: 0.75 -> 0.25 | Improved |
| use MOTCOM instead of density to rank sequences | complexity ranking · MOT20 test | normalized Footrule Distance: 0.50 -> 0.00 | Improved |
| use MOTCOM instead of density or tracks | complexity ranking · MOT17+MOT20 test combined | normalized Footrule Distance: 0.70 -> 0.27 | Improved |
| use MOTCOM instead of density | complexity ranking, CenterTrack ground truth · MOTSynth | normalized Footrule Distance: 0.63 -> 0.26 | Improved |
| use MOTCOM instead of tracks | complexity ranking, CenterTrack ground truth · MOTSynth | normalized Footrule Distance: 0.51 -> 0.26 | Improved |
| use MOTCOM instead of density | complexity ranking, CenterTrack ground truth · MOT train sequences | normalized Footrule Distance: 0.77 -> 0.40 | Improved |
| use MOTCOM instead of tracks | complexity ranking, CenterTrack ground truth · MOT test sequences | normalized Footrule Distance: 0.50 -> 0.43 | Improved |
| combine sub-metrics with arithmetic mean instead of quadratic, geometric or harmonic mean (ablation) | MOTCOM · MOT17+MOT20 test | Spearman correlation with HOTA | Improved |
Closest papers by meaning
- MOTS: Multi-Object Tracking and Segmentation (CVPR 2019 · similarity 0.89)
- More Than Meets the Eye: Enhancing Multi-Object Tracking Even with Prolonged Occlusions (ICML 2025 · similarity 0.89)
- HOTA: A Higher Order Metric for Evaluating Multi-object Tracking (IJCV 2020 · similarity 0.89)
- OVT-B: A New Large-Scale Benchmark for Open-Vocabulary Multi-Object Tracking (NeurIPS 2024 · similarity 0.88)
- DeconfuseTrack: Dealing with Confusion for Multi-Object Tracking (CVPR 2024 · similarity 0.87)
- Hybrid-SORT: Weak Cues Matter for Online Multi-Object Tracking (AAAI 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.
