Research map / Detection and segmentation / Weakly supervised detection and segmentation
Pseudo-labeling for semi-supervised detection
Train detectors on unlabeled images with teacher-student pseudo labels and distillation.
76 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2016–2026) · an approach in Weakly supervised detection and segmentation, keyword cluster “semi supervised · pseudo · unlabeled”
State History Benchmarks Tried Baselines Checklist Preprints Who Export
Where it stands
Few recent reports None of the tracked benchmarks has a report from 2025–26.
23 papers in 2023–24 and 17 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic grew, from 20.0% to 23.6%.
No longer reported by papers here: COCO-additional (AP50:95, last in 2022); MS-COCO, 10% labeled (mAP, last in 2023).
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 · Large Scale Semi-Supervised Object Detection Using Visual and Semantic Knowledge Transfer (CVPR 2016 · 160 citations)
- 2018 · Data Distillation: Towards Omni-Supervised Learning (CVPR 2018 · 437 citations)
- 2019 · Consistency-based Semi-supervised Learning for Object detection (NeurIPS 2019 · 282 citations)
- 2020 · Weak Supervision for Fake News Detection via Reinforcement Learning (AAAI 2020 · 152 citations)
- 2021 · Instant-Teaching: An End-to-End Semi-Supervised Object Detection Framework (CVPR 2021 · 222 citations)
- 2022 · Unbiased Teacher v2: Semi-Supervised Object Detection for Anchor-Free and Anchor-Based Detectors (CVPR 2022 · 142 citations)
- 2023 · Consistent-Teacher: Towards Reducing Inconsistent Pseudo-Targets in Semi-Supervised Object Detection (CVPR 2023 · 119 citations)
- 2024 · Sparse Semi-DETR: Sparse Learnable Queries for Semi-Supervised Object Detection (CVPR 2024 · 45 citations)
- 2025 · Multi-clue Consistency Learning to Bridge Gaps Between General and Oriented Object in Semi-supervised Detection (AAAI 2025 · 7 citations)
- 2026 · SeViL: Semi-supervised Vision-Language Learning with Text Prompt Guiding for Moving Infrared Small Target Detection (AAAI 2026 · 3 citations)
Earliest papers here
- Large Scale Semi-Supervised Object Detection Using Visual and Semantic Knowledge Transfer (CVPR 2016 · 160 citations)
- Data Distillation: Towards Omni-Supervised Learning (CVPR 2018 · 437 citations)
- Lagrangian Constrained Community Detection (AAAI 2018 · 7 citations)
Benchmark records
The best number each year that beat every earlier one, among papers of this approach reporting their own method. Numbers are extracted by AI from each paper's text; check the paper before citing.
COCO-additional · AP50:95 (higher is better)
4 papers, 2021–2022 · not reported since 2022
| Year | Best so far | Set by |
|---|---|---|
| 2021 | 41.3 | Unbiased Teacher for Semi-Supervised Object Detection (ICLR 2021 · 28 citations) |
| 2022 | 45.3 | Label Matching Semi-Supervised Object Detection (CVPR 2022 · 93 citations) |
MS-COCO, 10% labeled · mAP (higher is better)
4 papers, 2021–2023 · not reported since 2023
| Year | Best so far | Set by |
|---|---|---|
| 2021 | 29.8 | Instant-Teaching: An End-to-End Semi-Supervised Object Detection Framework (CVPR 2021 · 222 citations) |
| 2022 | 36.2 | Dense Learning Based Semi-Supervised Object Detection (CVPR 2022 · 85 citations) |
| 2023 | 37.2 | Adapting Object Size Variance and Class Imbalance for Semi-supervised Object Detection (AAAI 2023 · 13 citations) |
What has been tried
62 papers here report 764 measured changes (a component added, removed or swapped, compared with a baseline): helped 75%, no clear effect 4%, hurt 18%, 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 Query Refinement | COCO-Partial 10% labels · mAP | 43.5 -> 43.8 | Helped | Sparse Semi-DETR: Sparse Learnable Queries for Semi-Supervised Object Detection (CVPR 2024 · 45 citations) |
| Add the self-boosting teacher to the teacher-student framework | KITTI val split, Car Easy · 3D AP | 89.5 -> 93.8 | Helped | HINTED: Hard Instance Enhanced Detector with Mixed-Density Feature Fusion for Sparsely-Supervised 3D Object Detection (CVPR 2024 · 29 citations) |
| Add time-flip and horizontal-flip test-time augmentation | Gen1 WSOD 1% labels · mAP (%) | 35.2 -> 38.2 | Helped | LEOD: Label-Efficient Object Detection for Event Cameras (CVPR 2024 · 18 citations) |
| Add pseudo supervision on point-annotated images | DOTA-v1.0, 5% annotated · mAP | 35.90 -> 43.91 | Helped | Relational Matching for Weakly Semi-Supervised Oriented Object Detection (CVPR 2024 · 12 citations) |
| Use class-probability sampling rather than uniform sampling | ScanNet 10% · mAP@0.25 | 48.32 -> 50.63 | Helped | Dual-Perspective Knowledge Enrichment for Semi-supervised 3D Object Detection (AAAI 2024 · 11 citations) |
| Remove the refinement model | Waymo, 1% labeled + 4% unlabeled, Vehicle L1 · mAP | 54.4 -> 52.4 | Hurt | A-Teacher: Asymmetric Network for 3D Semi-Supervised Object Detection (CVPR 2024 · 8 citations) |
| Add GCA | DOTA-v1.5-Partial 30% · mAP(%) | 59.65 -> 60.62 | Helped | Multi-clue Consistency Learning to Bridge Gaps Between General and Oriented Object in Semi-supervised Detection (AAAI 2025 · 7 citations) |
| Add APM to teacher-student framework with a 1×1 convolution decoder | COD10K-Test · S-measure | .626 -> .633 | Helped | UCOD-DPL: Unsupervised Camouflaged Object Detection via Dynamic Pseudo-label Learning (CVPR 2025 · 5 citations) |
| Add negative sample assignment | DOTA-v1.0 · mAP50 | 23.62 -> 44.75 | Helped | PointOBB-v2: Towards Simpler, Faster, and Stronger Single Point Supervised Oriented Object Detection (ICLR 2025 · 5 citations) |
| Use explicit rather than implicit positional query initialization | nuScenes val · mAP | 57.58 -> 68.67 (+11.09) | Helped | Leveraging Imagery Data with Spatial Point Prior for Weakly Semi-supervised 3D Object Detection (AAAI 2024 · 4 citations) |
| Add TE Encoder | DroneVehicle (Point-Supervision) · AP50 | 35.26 -> 36.18 | Helped | PointSR: Self-Regularized Point Supervision for Drone-View Object Detection (CVPR 2025 · 3 citations) |
| add prototypical metric learning | PASCAL VOC Split-4 Extreme · AP50 | 75.0 -> 75.7 | Helped | As Pseudo-Label Free as Possible: Leveraging Adaptive Feature Generation for Sparsely Annotated Object Detection (AAAI 2025 · 0 citations) |
Baseline numbers that disagree
The same baseline on the same benchmark and metric, reported more than 5% apart by different papers here. Training budgets, data splits and reimplementations differ, so state which number you compare with.
| Baseline | Benchmark | Metric | Reported | Papers |
|---|---|---|---|---|
| Unbiased Teacher | COCO-standard 0.5% | AP50:95 | 15.95 – 16.94 | 2 papers, e.g. CVPR 2022 and AAAI 2022 |
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
- DOTA-v1.5 · 2 of 7 papers
- DOTA-v1.0 · 2 of 7 papers
Baselines
- Unbiased Teacher · 2 of 7 papers
- PointOBB · 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 40 papers; from OpenAlex and the papers' affiliations).
- Xidian University · 3 papers
- Tencent · 3 papers
- Baidu · 3 papers
- Sun Yat-sen University · 2 papers
- Zhejiang Lab · 2 papers
- Shanghai Jiao Tong University · 2 papers
- Tsinghua University · 2 papers
- Zhejiang University · 2 papers
- Peking University · 2 papers
- Xiamen University · 2 papers
Use it
Download BibTeX of 40 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
- Consistency regularization for segmentation (25 papers)
- Semi-supervised instance segmentation (16 papers)
Elsewhere in the topic
- Weakly supervised semantic segmentation (116 papers)
- Weakly supervised detection (97 papers)
- Weakly supervised object localization (54 papers)
- CAM refinement and prototypes (52 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).