Research map / Detection and segmentation / Tracking
Lost and Found: Overcoming Detector Failures in Online Multi-Object Tracking
Lorenzo Vaquero, Yihong Xu, Xavier Alameda-Pineda, Victor M. Brea, Manuel Mucientes
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 BUSCA to online trackers | TransCenter · MOT17 test | HOTA: 56.6 -> 61.7 | Improved |
| add BUSCA to CenterTrack | CenterTrack · MOT20 test | IDF1: 36.6 -> 58.0 | Improved |
| add BUSCA to ByteTrack | ByteTrack · MOT17 test | MOTA: 78.9 -> 79.3 | Improved |
| add BUSCA to GHOST | GHOST · MOT17 test | IDF1: 77.1 -> 77.0 | No clear effect |
| add BUSCA to TransCenter on MOT17 test MOTA | TransCenter · MOT17 test | MOTA: 76.2 -> 76.2 | No clear effect |
| lower detection threshold to 0.01 before association (ablation) | ByteTrack · MOT17 val | MOTA: 76.5 -> 75.3 | Hurt |
| add extra Kalman IoU recovery association round (ablation) | ByteTrack · MOT17 val | MOTA: 76.5 -> 75.4 | Hurt |
| add Mixed IoU and appearance recovery round (ablation) | ByteTrack · MOT17 val | MOTA: 76.5 -> 76.6 | No clear effect |
| use only [Halluc.] learned token (ablation) | BUSCA · MOT17 val | MOTA: 76.5 -> 75.0 | Hurt |
| combine [Halluc.] and [Miss.] tokens (ablation) | BUSCA · MOT17 val | MOTA: 75.0 -> 76.5 | Improved |
| add spatiotemporal encoding (STE) (ablation) | BUSCA · MOT17 val | MOTA: 76.5 -> 76.7 | Improved |
| add contextual proposals from neighbors (ablation) | BUSCA · MOT17 val | MOTA: 76.7 -> 77.1 | 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.
- ByteTrack* — ByteTrack: Multi-Object Tracking by Associating Every Detection Box (ECCV 2022) · better on 4 of 4 shared results
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
- Focusing on Tracks for Online Multi-Object Tracking (CVPR 2025 · similarity 0.89)
- Beyond Local Search: Tracking Objects Everywhere With Instance-Specific Proposals (CVPR 2016 · similarity 0.89)
- Tracking Without Bells and Whistles (ICCV 2019 · similarity 0.88)
- Track-On: Transformer-based Online Point Tracking with Memory (ICLR 2025 · similarity 0.87)
- Hybrid-SORT: Weak Cues Matter for Online Multi-Object Tracking (AAAI 2024 · similarity 0.87)
- DiffusionTrack: Diffusion Model for 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.
