Research map / Detection and segmentation / Tracking
Unified transformer tracking
Unify single and multi-object tracking with transformers and tracklet matching.
25 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2016–2026) · an approach in Tracking, keyword cluster “mot · association · object tracking”
State History Tried Checklist Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
3 papers in 2023–24 and 9 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic grew, from 4.5% to 11.2%.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2016 · Virtual Worlds as Proxy for Multi-Object Tracking Analysis (CVPR 2016 · 297 citations)
- 2017 · Visual Object Tracking for Unmanned Aerial Vehicles: A Benchmark and New Motion Models (AAAI 2017 · 410 citations)
- 2018 · Towards Dense Object Tracking in a 2D Honeybee Hive (CVPR 2018 · 42 citations)
- 2019 · CityFlow: A City-Scale Benchmark for Multi-Target Multi-Camera Vehicle Tracking and Re-Identification (CVPR 2019 · 464 citations)
- 2020 · Seeing Around Street Corners: Non-Line-of-Sight Detection and Tracking In-the-Wild Using Doppler Radar (CVPR 2020 · 148 citations)
- 2021 · Detection, Tracking, and Counting Meets Drones in Crowds: A Benchmark (CVPR 2021 · 137 citations)
- 2022 · Visible-Thermal UAV Tracking: A Large-Scale Benchmark and New Baseline (CVPR 2022 · 269 citations)
- 2024 · Multi-Object Tracking in the Dark (CVPR 2024 · 27 citations)
- 2025 · SUTrack: Towards Simple and Unified Single Object Tracking (AAAI 2025 · 50 citations)
- 2026 · Breaking Smooth-Motion Assumptions: A UAV Benchmark for Multi-Object Tracking in Complex and Adverse Conditions (CVPR 2026 · 0 citations)
Earliest papers here
- Virtual Worlds as Proxy for Multi-Object Tracking Analysis (CVPR 2016 · 297 citations)
- Visual Object Tracking for Unmanned Aerial Vehicles: A Benchmark and New Motion Models (AAAI 2017 · 410 citations)
- Object Detection in Videos With Tubelet Proposal Networks (CVPR 2017 · 161 citations)
What has been tried
22 papers here report 229 measured changes (a component added, removed or swapped, compared with a baseline): helped 69%, no clear effect 5%, hurt 24%, mixed 2%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| train separate single-task models instead of one multi-task model | five SOT task benchmarks · average score change | -7.8 | Hurt | SUTrack: Towards Simple and Unified Single Object Tracking (AAAI 2025 · 50 citations) |
| Add degradation suppression learning (DSL) without ALD | LMOT test · HOTA | 28.1 -> 29.0 | Helped | Multi-Object Tracking in the Dark (CVPR 2024 · 27 citations) |
| Use 2 sequential search frames instead of 1 | MUST test · AUC | 54.6 -> 59.7 | Helped | MUST: The First Dataset and Unified Framework for Multispectral UAV Single Object Tracking (CVPR 2025 · 14 citations) |
| add Track Query Group and Short-Term Memory Aggregator | MOT17 validation · MOTA | 67.4 -> 72.1 | Helped | TGFormer: Transformer with Track Query Group for Multi-Object Tracking (AAAI 2025 · 6 citations) |
| add PDR | SfM, cylindrical can · circle-fit R2 | 0.94 -> 0.9991 | Helped | Projecting Trackable Thermal Patterns for Dynamic Computer Vision (CVPR 2024 · 5 citations) |
| Add appearance-based Re-ID to BoT-SORT | BuckTales MOT · IDF1 | 0.9214 -> 0.9451 | Helped | BuckTales: A multi-UAV dataset for multi-object tracking and re-identification of wild antelopes (NeurIPS 2024 · 4 citations) |
| add semantic consistency and spatial proximity constraints | VOST val · J | 25.7 -> 50.9 | Helped | Tracking and Understanding Object Transformations (NeurIPS 2025 · 2 citations) |
| add Spectral 3D-Stem instead of using a 2D stem | MMOT MSI · HOTA | 40.3 -> 40.5 | Helped | MMOT: The First Challenging Benchmark for Drone-based Multispectral Multi-Object Tracking (NeurIPS 2025 · 1 citation) |
| Pretrain on the full colorspace | FeatureTracker · performance | Helped | Tracking objects that change in appearance with phase synchrony (ICLR 2025 · 0 citations) |
What reviewers will expect
Datasets and baselines named in the 7 papers of this approach from 2025–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Datasets
- GOT-10k · 2 of 7 papers
Baselines
- ODTrack · 2 of 7 papers
- MOTR · 2 of 7 papers
- MeMOTR · 2 of 7 papers
- MOTRv2 · 2 of 7 papers
- ByteTrack · 2 of 7 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 12 papers; from OpenAlex and the papers' affiliations).
- Beijing Institute of Technology · 3 papers
Use it
Download BibTeX of 25 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Association cues for online MOT (60 papers)
- Memory and correlation-based MOT (52 papers)
Elsewhere in the topic
- Transformer-based video object tracking (74 papers)
- Foundation-model and correlation-filter tracking (66 papers)
- Siamese network tracking (34 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).