Research map / Detection and segmentation / Open-set and few-shot detection
Few-shot object detection
Detect novel classes from few examples with contrastive proposals and incremental learning.
67 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2017–2026) · an approach in Open-set and few-shot detection, keyword cluster “zero · shot object · shot segmentation”
State History Benchmarks Tried Checklist Preprints Who Export
Where it stands
Quiet None of the tracked benchmarks has a report from 2025–26.
23 papers in 2023–24 and 11 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic shrank, from 3.9% to 2.0%.
No longer reported by papers here: MSCOCO, 10-shot (AP, last in 2024).
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
- 2017 · Few-Shot Object Recognition From Machine-Labeled Web Images (CVPR 2017 · 62 citations)
- 2018 · Hierarchical Novelty Detection for Visual Object Recognition (CVPR 2018 · 80 citations)
- 2019 · RepMet: Representative-Based Metric Learning for Classification and Few-Shot Object Detection (CVPR 2019 · 408 citations)
- 2020 · Few-Shot Object Detection With Attention-RPN and Multi-Relation Detector (CVPR 2020 · 638 citations)
- 2021 · FSCE: Few-Shot Object Detection via Contrastive Proposal Encoding (CVPR 2021 · 553 citations)
- 2022 · Grounded Language-Image Pre-Training (CVPR 2022 · 1,055 citations)
- 2023 · Few-Shot Object Detection via Variational Feature Aggregation (AAAI 2023 · 139 citations)
- 2024 · Few-Shot Object Detection with Foundation Models (CVPR 2024 · 64 citations)
- 2025 · CP-DETR: Concept Prompt Guide DETR Toward Stronger Universal Object Detection (AAAI 2025 · 11 citations)
- 2026 · FSOD-VFM: Few-Shot Object Detection with Vision Foundation Models and Graph Diffusion (ICLR 2026 · 0 citations)
Earliest papers here
- Few-Shot Object Recognition From Machine-Labeled Web Images (CVPR 2017 · 62 citations)
- Hierarchical Novelty Detection for Visual Object Recognition (CVPR 2018 · 80 citations)
- Learning to Promote Saliency Detectors (CVPR 2018 · 73 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.
MSCOCO, 10-shot · AP (higher is better)
4 papers, 2021–2024 · not reported since 2024
| Year | Best so far | Set by |
|---|---|---|
| 2021 | 32.1 | Generalized Few-Shot Object Detection Without Forgetting (CVPR 2021 · 195 citations) |
| 2024 | 40 | Few-Shot Object Detection with Foundation Models (CVPR 2024 · 64 citations) |
What has been tried
62 papers here report 725 measured changes (a component added, removed or swapped, compared with a baseline): helped 75%, no clear effect 5%, 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 |
|---|---|---|---|---|
| Use a larger DINOv2 backbone, ViT-L instead of ViT-B | MSCOCO, 10-shot · AP | 37.9 -> 40.0 | Helped | Few-Shot Object Detection with Foundation Models (CVPR 2024 · 64 citations) |
| replace grid-point prompts with class-agnostic heatmap localization | FSC-147 test · AP | 41.66 -> 43.53 | Helped | Point Segment and Count: A Generalized Framework for Object Counting (CVPR 2024 · 34 citations) |
| Add CutMix | PASCAL VOC, 3-split mean · nAP50 | 51.6 -> 52.8 | Helped | SNIDA: Unlocking Few-Shot Object Detection with Non-linear Semantic Decoupling Augmentation (CVPR 2024 · 30 citations) |
| Replace semantic class embeddings with visual description embeddings | DIOR ZSD · mAP | 6.1 -> 7.1 | Helped | Zero-Shot Aerial Object Detection with Visual Description Regularization (AAAI 2024 · 22 citations) |
| use prompt-based cluster selection | FSC147mul · MAE | 12.86 -> 6.46 | Helped | DAVE - A Detect-and-Verify Paradigm for Low-Shot Counting (CVPR 2024 · 19 citations) |
| Use the prompt visual hybrid encoder instead of a DINO encoder | LVIS minival, zero-shot · AP | 42.2 -> 44.3 | Helped | CP-DETR: Concept Prompt Guide DETR Toward Stronger Universal Object Detection (AAAI 2025 · 11 citations) |
| Add the upsampling module U | Animal pose, 1-shot novel keypoints · PCK | 39.82 -> 41.71 | Helped | Detect Any Keypoints: An Efficient Light-Weight Few-Shot Keypoint Detector (AAAI 2024 · 11 citations) |
| Add Rep+ (differentiated branch learning rates and reparameterization) | ODinW-13 · Avg AP | 58.12 -> 59.98 | Helped | Zero-shot Generalizable Incremental Learning for Vision-Language Object Detection (NeurIPS 2024 · 6 citations) |
| Use objectness scores rather than random soft labels for unknown pseudo-labels | M-OWODB Task 1 · U-Recall | 10.2 -> 23.6 | Helped | Open-World Objectness Modeling Unifies Novel Object Detection (CVPR 2025 · 5 citations) |
| Replace the proposed dense detection loss with standard unit-Gaussian loss | FSCD147 validation · MAE | 9.52 -> 12.79 | Hurt | A Novel Unified Architecture for Low-Shot Counting by Detection and Segmentation (NeurIPS 2024 · 5 citations) |
| Add ACR (auto-correlation representation) | KAIST-E unseen · F1 | 88.39 -> 92.42 | Helped | Spotting the Unseen: Reciprocal Consensus Network Guided by Visual Archetypes (AAAI 2024 · 4 citations) |
| add base-class suppression (BcS) module | COCO split-2 novel classes, 1-shot · AP50 | +1.5 AP50 (28.5 to 30.0) | Helped | Exploring Base-Class Suppression with Prior Guidance for Bias-Free One-Shot Object Detection (AAAI 2024 · 3 citations) |
What reviewers will expect
Datasets and baselines named in the 8 papers of this approach from 2025–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Datasets
- COCO · 2 of 8 papers
Baselines
- DeFRCN · 2 of 8 papers
- TFA · 2 of 8 papers
- MPSR · 2 of 8 papers
- CLIP · 2 of 8 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 34 papers; from OpenAlex and the papers' affiliations).
- Sichuan University · 3 papers
- Peking University · 3 papers
- Chinese Academy of Sciences · 2 papers
- University of Chinese Academy of Sciences · 2 papers
- University of Science and Technology of China · 2 papers
- Tencent · 2 papers
- Australian National University · 2 papers
- Data61-CSIRO · 2 papers
- Columbia University · 2 papers
- Meta · 2 papers
Use it
Download BibTeX of 39 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
- Few-shot segmentation (64 papers)
- Zero-shot segmentation and detection (49 papers)
- Prompted SAM segmentation (39 papers)
Elsewhere in the topic
- Object-level representation learning (219 papers)
- Interactive and referring segmentation (161 papers)
- Open-set and unknown object detection (160 papers)
- Amodal and open-world instance segmentation (137 papers)
- Industrial anomaly detection (117 papers)
- Unsupervised object and part discovery (114 papers)
- Domain-adaptive object detection (103 papers)
- Open-vocabulary object detection (96 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).