Research map / Continual learning / Domain and prompt-based
Lifelong person re-identification
Accumulates and consolidates knowledge across domains for person re-identification with rehearsal or distillation.
14 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2021–2026) · an approach in Domain and prompt-based, keyword cluster “prompts · person · prompting”
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.
4 papers in 2023–24 and 6 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic grew, from 9.1% to 12.2%.
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
- 2021 · Lifelong Person Re-Identification via Adaptive Knowledge Accumulation (CVPR 2021 · 99 citations)
- 2022 · Lifelong Unsupervised Domain Adaptive Person Re-Identification With Coordinated Anti-Forgetting and Adaptation (CVPR 2022 · 51 citations)
- 2023 · Lifelong Person Re-identification via Knowledge Refreshing and Consolidation (AAAI 2023 · 38 citations)
- 2024 · LSTKC: Long Short-Term Knowledge Consolidation for Lifelong Person Re-identification (AAAI 2024 · 26 citations)
- 2025 · DASK: Distribution Rehearsing via Adaptive Style Kernel Learning for Exemplar-Free Lifelong Person Re-Identification (AAAI 2025 · 9 citations)
- 2026 · Prompt-Anchored Vision-Text Distillation for Lifelong Person Re-identification (CVPR 2026 · 0 citations)
Earliest papers here
- Lifelong Person Re-Identification via Adaptive Knowledge Accumulation (CVPR 2021 · 99 citations)
- Generalising without Forgetting for Lifelong Person Re-Identification (AAAI 2021 · 48 citations)
- Lifelong Unsupervised Domain Adaptive Person Re-Identification With Coordinated Anti-Forgetting and Adaptation (CVPR 2022 · 51 citations)
What has been tried
10 papers here report 118 measured changes (a component added, removed or swapped, compared with a baseline): helped 69%, no clear effect 4%, hurt 24%, mixed 3%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| add R-STKT relation rectification and transfer | LReID benchmark, training order-1 · mAP | 40.4 -> 46.1 | Helped | LSTKC: Long Short-Term Knowledge Consolidation for Lifelong Person Re-identification (AAAI 2024 · 26 citations) |
| Add IDM | Seen-Avg · mAP | 44.0 -> 46.1 | Helped | Distribution-aware Knowledge Prototyping for Non-exemplar Lifelong Person Re-identification (CVPR 2024 · 22 citations) |
| Use adaptive rather than fixed am | Order-1 RFL-ReID · mAP | 39.5 vs 38.2, 38.7, 39.0, 38.4, 38.9, 38.7 | Helped | Learning Continual Compatible Representation for Re-indexing Free Lifelong Person Re-identification (CVPR 2024 · 15 citations) |
| Add distribution rehearsal using DASK's adaptive transfer kernel rather than training on new data only | Seen-Avg · mAP | 48.7 -> 55.4 | Helped | DASK: Distribution Rehearsing via Adaptive Style Kernel Learning for Exemplar-Free Lifelong Person Re-Identification (AAAI 2025 · 9 citations) |
| Add DKT | CH-LReID Order-1 average · mAP | 41.7 -> 43.1 | Helped | DKC: Differentiated Knowledge Consolidation for Cloth-Hybrid Lifelong Person Re-identification (CVPR 2025 · 4 citations) |
What reviewers will expect
Datasets and baselines named in the 6 papers of this approach from 2024–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Datasets
- Market-1501 · 5 of 6 papers
- CUHK-SYSU · 5 of 6 papers
- DukeMTMC-reID · 5 of 6 papers
- MSMT17-V2 · 5 of 6 papers
- CUHK03 · 5 of 6 papers
- CUHK01 · 2 of 6 papers
- CUHK02 · 2 of 6 papers
- VIPeR · 2 of 6 papers
Baselines
- LwF · 5 of 6 papers
- CRL · 5 of 6 papers
- AKA · 5 of 6 papers
- PatchKD · 5 of 6 papers
- SPD · 4 of 6 papers
- JointTrain · 3 of 6 papers
- MEGE · 3 of 6 papers
- Finetune · 2 of 6 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 10 papers; from OpenAlex and the papers' affiliations).
- Peking University · 5 papers
- Huazhong University of Science and Technology · 2 papers
Use it
Download BibTeX of 14 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
- Prompt-based continual learning (26 papers)
Elsewhere in the topic
- Domain-incremental learning with memory (49 papers)
- Continual test-time adaptation (25 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).