Pith. sign in

REVIEW 2 major objections 5 minor 1 cited by

Few-Shot Object Detection with Attention-RPN and Multi-Relation Detector

T0 review · 2 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read After one training pass on 800 object categories, a few-shot detector can localize objects in 200 categories it never saw, using only a few support images per category, with no fine-tuning.

desk verdict Solid few-shot detection paper with a genuinely useful dataset; the 'unseen category' no-fine-tune claim needs a leakage check on the FSOD split, but the method's gains hold even without FSOD. read the letter →

arxiv 1908.01998 v4 pith:LKR52EIZ submitted 2019-08-06 cs.CV

Qi Fan , Wei Zhuo , Chi-Keung Tang , Yu-Wing Tai This is my paper · ORCID
classification cs.CV
keywords few-shotobjectdetectionunseencategoriesattentionRPNmulti-relationdetectorcontrastivetrainingFSODdatasetsupport-querymatchinggeneralization
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper claims that object detection can be made to work for categories a detector has never been trained on: one network, trained once, can localize a novel object after seeing only a few annotated examples of it, with no retraining or fine-tuning. To support that claim the authors built the FSOD dataset, 1000 categories drawn from ImageNet and Open Images with an 800/200 training/test split chosen so that the 200 test categories are semantically distant from training ones. Trained on FSOD, their model reaches 41.7 AP50 on ImageNet Detection (50-way 5-shot) and 16.6 AP / 31.3 AP50 on MS COCO minival (20-way 10-shot) without any fine-tuning, beating prior few-shot detectors that are fine-tuned on the novel categories. If the claim is right, it means few-shot detection can be decoupled from per-task retraining, and that the number of training categories, not the number of training images, is the resource that buys generalization to unseen objects.

What carries the argument

Four components carry the argument. The Attention-RPN injects the support category into proposal generation: the support feature is average-pooled to a $1\times1$ vector and used as a depth-wise cross-correlation kernel over the query feature map, so proposals are scored by similarity to the supported object rather than by generic objectness. The Multi-Relation Detector scores each query proposal against the support with three complementary heads: a global-relation head that matches pooled embeddings, a local-correlation head that computes pixel-wise depth-wise similarity, and a patch-relation head that learns a non-linear patch metric; their scores are summed. The 2-way contrastive training strategy builds triplets of query, positive-support, and negative-support images, balancing foreground-positive, background-positive, and negative-support pairs at a 1:2:1 ratio, so the network learns to reject background boxes and wrong categories while matching the target. The FSOD dataset supplies the diversity and the split: 1000 categories merged from ImageNet and Open Images, with the 200 test categories chosen by shortest-path distance in the WordNet is-a taxonomy from the 800 training categories, which is what makes an 'unseen category' evaluation possible at all.

What would settle it

Take a pre-trained network, compute per-category feature centroids for the FSOD training and test sets, and measure nearest-neighbour distances between each of the 200 test categories and the 800 training categories; if many test categories are closer to some training category than their own examples are to each other (for instance 'melon' close to the trained 'watermelon'), the no-fine-tuning gains could come from visual overlap rather than transfer to unseen categories. A second check: re-run the Table 4 and Table 5 evaluations after deleting from the query images any objects whose categories also appear in the training set, and see whether the reported AP50 margins over the fine-tuned baselines survive.

Watch

Extended reading notes

Core claim

The paper's central claim is that a few-shot object detector can learn a general support-to-query matching relationship, and that this relationship transfers to categories withheld from training. The evidence is the no-fine-tuning column of the benchmark tables: the model trained only on the FSOD training split detects the 200 held-out FSOD test categories at 27.5 AP50, and reaches 41.7 AP50 on ImageNet Detection 50-way 5-shot and 16.6 AP / 31.3 AP50 on MS COCO minival 20-way 10-shot, while the comparison methods (LSTD, RepMet, Feature Reweighting, Meta R-CNN) all fine-tune on the novel categories and score lower. The authors also argue that category diversity drives the transfer: with roughly a quarter of COCO's images, FSOD's 800 training categories outperform COCO's 80 categories on the held-out set, and performance rises monotonically as training categories are added at fixed image count. The stated conclusion is that learning to match support and query across many categories generalizes better than learning category-specific embeddings.

Load-bearing premise

The load-bearing premise is that the 200 test categories are genuinely new to the model: the training/test split was made by measuring word meanings and merging labels by hand, but the paper never checks whether the test categories' images, objects, or scene contexts actually look like, or co-occur with, the ones in the 800 training categories.

Editorial extensions

If this is right

  • Deploying a detector on a new category reduces to supplying a few support images: the same trained weights serve every new category, so adding a category costs only a few annotations, not a training run.
  • Dataset design for few-shot detection should prioritize category breadth: with similar image counts, 800 training categories beat 80 by a wide margin on held-out categories (64.7 vs 49.1 AP50), so collecting diverse categories is more valuable than collecting many images per category.
  • Filtering at the proposal stage is measurable and load-bearing: the attention RPN raises proposal recall (0.9130 vs 0.8804 at top-100 over 0.5 IoU) and average best overlap (0.7282 vs 0.7127), which is what makes the downstream matching detector feasible.
  • One negative support category is enough for contrastive training: 2-way 5-shot training (67.5 AP50) beats 5-way 5-shot training (66.9 AP50), so the contrastive signal saturates quickly.
  • Fine-tuning becomes optional rather than mandatory: the FSOD-trained model without fine-tuning already beats all compared methods, and fine-tuning on the target set adds a further gain, so fine-tuning can be treated as a refinement.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A direct check of the paper's 'unseen categories' premise would be a feature-level nearest-neighbour analysis between the 200 held-out and 800 training categories; if visual leakage exists (melon vs watermelon, squid vs shrimp), part of the reported transfer could be low-level feature overlap, not generalization to genuinely new categories.
  • The support-conditioned proposal mechanism is the same depth-wise cross-correlation used in Siamese visual tracking, a line the paper itself cites; a natural extension is video few-shot detection, where the support is an annotated first frame and the tracking-style matching propagates detection to later frames.
  • The 1:2:1 sampling ratio among positive, background, and negative-support pairs is a design choice the paper does not sweep; varying it or mining hard negatives among background proposals could bound how much of the gain is due to contrastive training specifically.
  • The no-fine-tuning claim is demonstrated at 5-shot and 10-shot with 50 and 20 novel categories; pushing to 1-shot and to larger held-out sets would locate where proposal quality and contrastive signal degrade enough that fine-tuning becomes necessary.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. The paper proposes a few-shot object detection framework combining an Attention-RPN, a Multi-Relation Detector, and a 2-way contrastive training strategy, together with a new 1000-category FSOD dataset assembled from ImageNet and Open Images. The central claim is that after training on the FSOD training categories, the model can detect objects of unseen categories without any fine-tuning, and that it outperforms prior few-shot detectors on ImageNet Detection (50-way 5-shot) and MS COCO minival (20-way 10-shot). The paper also includes ablations on relation-head design, training strategy, RPN attention, and dataset size, plus applications to wild car and penguin detection.

Significance. If the no-fine-tuning generalization claim is sound, this is a substantial advance for few-shot object detection, and the public release of a 1000-category dataset is a valuable community resource. The paper deserves credit for including same-training-set comparisons with prior methods (e.g., the COCO-trained rows in Tables 4 and 5), for reporting the FSOD dataset statistics, and for providing a detailed class split in Appendix D. However, the headline claim currently rests on the unverified assumption that the evaluation categories are truly absent from FSOD training, and several design choices are supported only by ablations without variance estimates. These issues materially affect the strength of the conclusions.

major comments (2)
  1. [Section 3 and Appendix D; Tables 4 and 5]
  2. [Section 4.3, Table 3, and Appendix B.2]
minor comments (5)
  1. [Section 3]
  2. [Section 4.2.1, Eq. (1)]
  3. [References]
  4. [Table 7]
  5. [Appendix B.2]

Circularity Check

1 steps flagged · score 6.0 of 10

The 'no fine-tune unseen category' results in Tables 4/5 are undermined by Appendix D: the FSOD training split released/used in experiments contains the very PASCAL VOC categories that Table 5 calls novel, so the prediction reduces to evaluating on seen categories.

  1. fitted input called prediction [Section 5.2.2, Table 5 footnotes; Appendix D FSOD Dataset Class Split]
    "† means that the testing categories are removed from FSOD training dataset. ... we set the 20 categories included in PASCAL VOC as novel categories for evaluation ... Appendix D: FSOD Dataset Class Split. Here we describe the training/testing class split in our proposed FSOD Dataset. This split was used in our experiments."

    Table 5 reports 'Ours FSOD† no fine-tune' with 16.6 AP / 31.3 AP50 on the 20 PASCAL VOC categories, and the dagger states that those testing categories are removed from FSOD training. However, Appendix D, which says 'This split was used in our experiments,' lists in the FSOD training split the very categories that PASCAL VOC defines as novel: 'cat' (in 'watch, cat, ladder'), 'horse' (in 'horse, duck, salad'), 'bus' (in 'doll, bus, flying disc'), 'train' (in 'starfish, train, belt'), and 'truck' (in 'pasta, truck, blue jay'), along with 'sheep', 'cow'/'cattle', 'airplane', and other close entries. Consequently, the no-fine-tune FSOD model is not being evaluated on categories absent from its training inputs; it is being evaluated on categories that appear in the released training split.

full rationale

The paper's architecture and ablation studies (Attention-RPN, multi-relation detector, 2-way contrastive training) are self-contained and are compared against external baselines under shared protocols; those parts do not exhibit circular derivation. The central cross-dataset 'unseen category' claim, however, depends on the FSOD† training split truly excluding the evaluation categories. The manuscript contradicts itself: Table 5 states that the testing categories are removed, while Appendix D — explicitly labeled as the split used in the experiments — contains the PASCAL VOC novel categories (cat, dog, horse, bus, train, truck, sheep, airplane, etc.) in the FSOD training list. If the dagger-removed split differs from the released Appendix D split, its exact contents are unspecified and the load-bearing 'unseen' condition is unverified; if the Appendix D split is the one used, the no-fine-tune results are fitted-input-called-prediction on seen categories. This is a partial circularity affecting the paper's headline claim, though it does not invalidate the method's architectural contributions or the internal FSOD test-set experiments, whose 200 test categories are disjoint from the 800 training categories by construction.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

No physical constants are fitted; the model's weights are learned from data. The free parameters listed are hand-set hyperparameters whose values affect the reported results. The axioms are the unproved transfer and evaluation assumptions on which the empirical claims rest.

free parameters (6)
  • contrastive pair ratio = 1:2:1 (foreground-positive : background-positive : negative)
    Chosen by hand in Section 4.3 to balance matching pairs; performance depends on it, and no sensitivity analysis is provided.
  • support kernel size S = 1
    Section 4.2.1: 'We find that a kernel size of S = 1 performs well'; hand-selected, with no ablation shown for other S values.
  • training schedule = lr 0.002 for 56000 iterations, lr 0.0002 for 4000 iterations
    Section 5.1 hand-set; the authors note that more iterations damage performance, but no grid search or sensitivity analysis is reported.
  • support image size = 320x320
    Section 5.1; hand-set for support preprocessing.
  • query image scale = shorter side 600, longer side capped at 1000
    Section 5.1; standard Faster R-CNN scale but still a hand-set input configuration.
  • frozen backbone blocks = ResNet50 blocks 1-3 fixed
    Section 5.1; a choice to avoid overfitting, not systematically justified by experiments.
assumptions (5)
  • domain assumption Pretrained ResNet50 on ImageNet and COCO provides stable low-level features that transfer to few-shot detection.
    Section 5.1 states 'pre-training on ImageNet and MS COCO can provide stable low-level features'; this is assumed, not independently proven.
  • domain assumption Depth-wise cross-correlation with a global average-pooled support feature (S=1) encodes category-specific attention useful for suppressing non-support proposals.
    Section 4.2.1; the design is justified by consistency with [25] and empirical recall/ABO gains, but not by an independent mechanism.
  • domain assumption A single negative support category in 2-way contrastive training is sufficient to teach discrimination across all categories.
    Section 4.3: 5-way training did not beat 2-way; authors conclude 'only one negative support category suffices', but differences are small and no significance test is reported.
  • domain assumption The FSOD train/test split by WordNet semantic distance ensures the test categories are unseen and not semantically entangled with training categories.
    Section 3; the split uses shortest path in an is-a taxonomy and manual merging, but no quantitative verification of label cleanliness or absence of visual overlap is given.
  • domain assumption The 600-episode evaluation protocol is a faithful proxy for standard detection performance.
    Appendix B.2; uses random episodes with 10 query images per category; no variance reported, and the protocol is borrowed from RepMet [61] rather than justified.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Few-Shot Object Detection with Attention-RPN and Multi-Relation Detector." pith.science (2026). https://pith.science/paper/LKR52EIZ

@misc{pith2026190801998,
  author       = {Pith},
  title        = {Pith review of: Few-Shot Object Detection with Attention-RPN and Multi-Relation Detector},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LKR52EIZ}},
  note         = {Machine review of arXiv:1908.01998}
}
read the original abstract

Conventional methods for object detection typically require a substantial amount of training data and preparing such high-quality training data is very labor-intensive. In this paper, we propose a novel few-shot object detection network that aims at detecting objects of unseen categories with only a few annotated examples. Central to our method are our Attention-RPN, Multi-Relation Detector and Contrastive Training strategy, which exploit the similarity between the few shot support set and query set to detect novel objects while suppressing false detection in the background. To train our network, we contribute a new dataset that contains 1000 categories of various objects with high-quality annotations. To the best of our knowledge, this is one of the first datasets specifically designed for few-shot object detection. Once our few-shot network is trained, it can detect objects of unseen categories without further training or fine-tuning. Our method is general and has a wide range of potential applications. We produce a new state-of-the-art performance on different datasets in the few-shot setting. The dataset link is https://github.com/fanq15/Few-Shot-Object-Detection-Dataset.

Figures

Figures reproduced from arXiv: 1908.01998 by the authors.

Figure 1
Figure 1. Given different objects as supports (top corners above), [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Dataset label tree. The ImageNet categories (red circles) are merged with Open Image categories (green circles) where the [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. The dataset statistics of FSOD. The category image num [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (10 more)
Figure 4
Figure 4. Figure 4: Network architecture. The query image and support image are processed by the weight-shared network. The attention RPN [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 5
Figure 5. Figure 5: Attention RPN. The support feature is average pooled to [PITH_FULL_IMAGE:figures/full_fig_p004_5.png]
Figure 6
Figure 6. Figure 6: Multi-Relation Detector. Different relation heads model [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]
Figure 7
Figure 7. Figure 7: The 2-way contrastive training triplet and different [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]
Figure 8
Figure 8. Figure 8: Qualitative 1-shot detection results of our approach on FSOD test set. Zoom in the figures for more visual details. [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 9
Figure 9. Figure 9: Our application results on the penguin dataset [ [PITH_FULL_IMAGE:figures/full_fig_p008_9.png]
Figure 10
Figure 10. Figure 10: Qualitative 1-shot object detection results on our test set. We visualize the bounding boxes with score larger than 0.8. [PITH_FULL_IMAGE:figures/full_fig_p012_10.png]
Figure 11
Figure 11. Figure 11: Qualitative results of our 1-shot object detection on test set. We visualize the bounding boxes with score larger than 0.8. [PITH_FULL_IMAGE:figures/full_fig_p013_11.png]
Figure 12
Figure 12. Figure 12: Qualitative results of our 5-shot car detection on Cityscapes. We visualize the bounding boxes with score larger than 0.8. The [PITH_FULL_IMAGE:figures/full_fig_p014_12.png]
Figure 13
Figure 13. Figure 13: Qualitative results of our 5-shot car detection on KITTI. We visualize the bounding boxes with score larger than 0.8. [PITH_FULL_IMAGE:figures/full_fig_p014_13.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Deep Learning in Automated Power Line Inspection: A Review

    cs.CV 2025-02 unverdicted novelty 1.0 of 10

    A review of deep learning for power line inspection, structured around component detection and fault diagnosis, with no novel experimental contributions.

Reference graph

Works this paper leans on

65 extracted references · 48 canonical work pages · cited by 1 Pith paper

  1. [58]

    Fully-convolutional siamese networks for object tracking

    Luca Bertinetto, Jack Valmadre, Joao F Henriques, Andrea Vedaldi, and Philip HS Torr. Fully-convolutional siamese networks for object tracking. In ECCV, 2016

    work page 2016

  2. [59]

    Class-agnostic counting

    Erika Lu, Weidi Xie, and Andrew Zisserman. Class-agnostic counting. In ACCV, 2018

    work page 2018

  3. [60]

    Siamrpn++: Evolution of siamese visual tracking with very deep networks

    Bo Li, Wei Wu, Qiang Wang, Fangyi Zhang, Junliang Xing, and Junjie Yan. Siamrpn++: Evolution of siamese visual tracking with very deep networks. In CVPR, 2019

    work page 2019

  4. [8]

    Learning to compare: Re- lation network for few-shot learning

    Flood Sung Yongxin Yang, Li Zhang, Tao Xiang, Philip HS Torr, and Timothy M Hospedales. Learning to compare: Re- lation network for few-shot learning. In CVPR, 2018

    work page 2018

  5. [10]

    Few-shot object detection via feature reweighting

    Bingyi Kang, Zhuang Liu, Xin Wang, Fisher Yu, Jiashi Feng, and Trevor Darrell. Few-shot object detection via feature reweighting. In ICCV, 2019

    work page 2019

  6. [1]

    Prototypical networks for few-shot learning

    Jake Snell, Kevin Swersky, and Richard Zemel. Prototypical networks for few-shot learning. In NeurIPS, 2017

    2017

  7. [2]

    Optimization as a model for few-shot learning

    Sachin Ravi and Hugo Larochelle. Optimization as a model for few-shot learning. In ICLR, 2017

    work page 2017

  8. [3]

    Meta-learning with memory-augmented neural networks

    Adam Santoro, Sergey Bartunov, Matthew Botvinick, Daan Wierstra, and Timothy Lillicrap. Meta-learning with memory-augmented neural networks. In ICML, 2016

    work page 2016

Show all 65 references
  1. [4]

    Matching networks for one shot learning

    Oriol Vinyals, Charles Blundell, Tim Lillicrap, Daan Wier- stra, et al. Matching networks for one shot learning. In NeurIPS, 2016

    2016

  2. [6]

    Memory matching networks for one-shot image recog- nition

    Qi Cai, Yingwei Pan, Ting Yao, Chenggang Yan, and Tao Mei. Memory matching networks for one-shot image recog- nition. In CVPR, 2018

    2018

  3. [7]

    Dynamic few-shot visual learning without forgetting

    Spyros Gidaris and Nikos Komodakis. Dynamic few-shot visual learning without forgetting. In CVPR, 2018

    2018

  4. [9]

    Lstd: A low-shot transfer detector for object detection

    Hao Chen, Yali Wang, Guoyou Wang, and Yu Qiao. Lstd: A low-shot transfer detector for object detection. In AAAI, 2018

    2018

  5. [11]

    Repmet: Representative-based metric learning for classification and few-shot object detection

    Leonid Karlinsky, Joseph Shtok, Sivan Harary, Eli Schwartz, Amit Aides, Rogerio Feris, Raja Giryes, and Alex M Bron- stein. Repmet: Representative-based metric learning for classification and few-shot object detection. In CVPR, 2019

    2019

  6. [12]

    Meta r-cnn : Towards general solver for instance-level low-shot learning

    Xiaopeng Yan, Ziliang Chen, Anni Xu, Xiaoxi Wang, Xi- aodan Liang, and Liang Lin. Meta r-cnn : Towards general solver for instance-level low-shot learning. In ICCV, 2019

    2019

  7. [13]

    Microsoft coco: Common objects in context

    Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Doll´ar, and C Lawrence Zitnick. Microsoft coco: Common objects in context. In ECCV, 2014

    2014

  8. [14]

    Histograms of oriented gra- dients for human detection

    Navneet Dalal and Bill Triggs. Histograms of oriented gra- dients for human detection. In CVPR, 2005

    2005

  9. [15]

    Object detection with discriminatively trained part-based models

    Pedro F Felzenszwalb, Ross B Girshick, David McAllester, and Deva Ramanan. Object detection with discriminatively trained part-based models. IEEE transactions on pattern analysis and machine intelligence, 32(9):1627–1645, 2010

    2010

  10. [16]

    Rapid object detection using a boosted cascade of simple features

    P VIODA. Rapid object detection using a boosted cascade of simple features. In CVPR, 2001

    2001

  11. [17]

    Imagenet classification with deep convolutional neural net- works

    Alex Krizhevsky, Ilya Sutskever, and Geoffrey E Hinton. Imagenet classification with deep convolutional neural net- works. In NeurIPS. 2012

    2012

  12. [18]

    You only look once: Unified, real-time object de- tection

    Joseph Redmon, Santosh Divvala, Ross Girshick, and Ali Farhadi. You only look once: Unified, real-time object de- tection. In CVPR, 2016

    2016

  13. [19]

    Yolo9000: better, faster, stronger

    Joseph Redmon and Ali Farhadi. Yolo9000: better, faster, stronger. In CVPR, 2017

    2017

  14. [20]

    Ssd: Single shot multibox detector

    Wei Liu, Dragomir Anguelov, Dumitru Erhan, Christian Szegedy, Scott Reed, Cheng-Yang Fu, and Alexander C Berg. Ssd: Single shot multibox detector. In ECCV, 2016. Figure 10. Qualitative 1-shot object detection results on our test set. We visualize the bounding boxes with score ...

    2016

  15. [21]

    Focal loss for dense object detection

    Tsung-Yi Lin, Priya Goyal, Ross Girshick, Kaiming He, and Piotr Doll´ar. Focal loss for dense object detection. In ICCV, 2017

    2017

  16. [22]

    Receptive field block net for accurate and fast object detection

    Songtao Liu, Di Huang, et al. Receptive field block net for accurate and fast object detection. In ECCV, 2018

    2018

  17. [23]

    Rich feature hierarchies for accurate object detection and semantic segmentation

    Ross Girshick, Jeff Donahue, Trevor Darrell, and Jitendra Malik. Rich feature hierarchies for accurate object detection and semantic segmentation. In CVPR, 2014

    2014

  18. [24]

    Fast r-cnn

    Ross Girshick. Fast r-cnn. In ICCV, 2015

    2015

  19. [25]

    Faster r-cnn: Towards real-time object detection with region proposal networks

    Shaoqing Ren, Kaiming He, Ross Girshick, and Jian Sun. Faster r-cnn: Towards real-time object detection with region proposal networks. In NeurIPS, 2015

    2015

  20. [26]

    Mask r-cnn

    Kaiming He, Georgia Gkioxari, Piotr Doll ´ar, and Ross Gir- shick. Mask r-cnn. In ICCV, 2017

    2017

  21. [27]

    Sniper: Efficient multi-scale training

    Bharat Singh, Mahyar Najibi, and Larry S Davis. Sniper: Efficient multi-scale training. In NeurIPS, 2018

    2018

  22. [28]

    Is learning the n-th thing any easier than learning the first? In NeurIPS, 1996

    Sebastian Thrun. Is learning the n-th thing any easier than learning the first? In NeurIPS, 1996

    1996

  23. [29]

    One-shot learning of object categories

    Li Fei-Fei, Rob Fergus, and Pietro Perona. One-shot learning of object categories. IEEE transactions on pattern analysis and machine intelligence, 28(4):594–611, 2006

    2006

  24. [30]

    One shot learning of simple visual con- cepts

    Brenden Lake, Ruslan Salakhutdinov, Jason Gross, and Joshua Tenenbaum. One shot learning of simple visual con- cepts. In Proceedings of the Annual Meeting of the Cognitive Science Society, volume 33, 2011

    2011

  25. [31]

    One-shot learning by inverting a compositional causal process

    Brenden M Lake, Ruslan R Salakhutdinov, and Josh Tenen- baum. One-shot learning by inverting a compositional causal process. In NeurIPS, 2013

    2013

  26. [32]

    Human-level concept learning through proba- bilistic program induction

    Brenden M Lake, Ruslan Salakhutdinov, and Joshua B Tenenbaum. Human-level concept learning through proba- bilistic program induction. Science, 350(6266):1332–1338, 2015

    2015

  27. [33]

    One shot learning via com- positions of meaningful patches

    Alex Wong and Alan L Yuille. One shot learning via com- positions of meaningful patches. In ICCV, 2015

    2015

  28. [34]

    Tadam: Task dependent adaptive metric for improved few-shot learning

    Boris Oreshkin, Pau Rodr ´ıguez L´opez, and Alexandre La- coste. Tadam: Task dependent adaptive metric for improved few-shot learning. In NeurIPS, 2018

    2018

  29. [35]

    Few-shot learning through an information retrieval lens

    Eleni Triantafillou, Richard Zemel, and Raquel Urtasun. Few-shot learning through an information retrieval lens. In NeurIPS, 2017

    2017

  30. [36]

    Low-shot visual recognition by shrinking and hallucinating features

    Bharath Hariharan and Ross Girshick. Low-shot visual recognition by shrinking and hallucinating features. In ICCV, 2017

    2017

  31. [37]

    Siamese neural networks for one-shot image recognition

    Gregory Koch, Richard Zemel, and Ruslan Salakhutdinov. Siamese neural networks for one-shot image recognition. In ICML Workshop, 2015

    2015

  32. [38]

    Meta networks

    Tsendsuren Munkhdalai and Hong Yu. Meta networks. In ICML, 2017

    2017

  33. [39]

    Rapid adaptation with conditionally shifted neurons

    Tsendsuren Munkhdalai, Xingdi Yuan, Soroush Mehri, and Adam Trischler. Rapid adaptation with conditionally shifted neurons. In ICML, 2018

    2018

  34. [40]

    Low-shot learning from imaginary data

    Yu-Xiong Wang, Ross Girshick, Martial Hebert, and Bharath Hariharan. Low-shot learning from imaginary data. In CVPR, 2018

    2018

  35. [41]

    Revisiting local descriptor based image-to-class measure for few-shot learning

    Wenbin Li, Lei Wang, Jinglin Xu, Jing Huo, Gao Yang, and Jiebo Luo. Revisiting local descriptor based image-to-class measure for few-shot learning. In CVPR, 2019

    2019

  36. [42]

    Dense classification and implanting for few-shot learning

    Yann Lifchitz, Yannis Avrithis, Sylvaine Picard, and Andrei Bursuc. Dense classification and implanting for few-shot learning. In CVPR, 2019

    2019

  37. [43]

    Yoo Jongmin Kim, Taesup Kim

    Sungwoong Kim Chang D. Yoo Jongmin Kim, Taesup Kim. Edge-labeling graph neural network for few-shot learning. In CVPR, 2019

    2019

  38. [44]

    Generating classifi- cation weights with gnn denoising autoencoders for few-shot learning

    Spyros Gidaris and Nikos Komodakis. Generating classifi- cation weights with gnn denoising autoencoders for few-shot learning. In CVPR, 2019

    2019

  39. [45]

    Finding task-relevant features for few- shot learning by category traversal

    Hongyang Li, David Eigen, Samuel Dodge, Matthew Zeiler, and Xiaogang Wang. Finding task-relevant features for few- shot learning by category traversal. In CVPR, 2019

    2019

  40. [46]

    Model- agnostic meta-learning for fast adaptation of deep networks

    Chelsea Finn, Pieter Abbeel, and Sergey Levine. Model- agnostic meta-learning for fast adaptation of deep networks. In ICML, 2017

    2017

  41. [47]

    Few-shot semantic segmen- tation with prototype learning

    Nanqing Dong and Eric P Xing. Few-shot semantic segmen- tation with prototype learning. In BMVC, 2018

    2018

  42. [48]

    Claudio Michaelis, Matthias Bethge, and Alexander S. Ecker. One-shot segmentation in clutter. In ICML, 2018

    2018

  43. [49]

    Tao Hu, Pengwan, Chiliang Zhang, Gang Yu, Yadong Mu, and Cees G. M. Snoek. Attention-based multi-context guid- ing for few-shot semantic segmentation. In AAAI, 2019

    2019

  44. [50]

    Liang-Yan Gui, Yu-Xiong Wang, Deva Ramanan, and Jos ´e M. F. Moura. Few-shot human motion prediction via meta- learning. In ECCV, 2018

    2018

  45. [51]

    Few-example object detection with model communi- cation

    Xuanyi Dong, Liang Zheng, Fan Ma, Yi Yang, and Deyu Meng. Few-example object detection with model communi- cation. IEEE transactions on pattern analysis and machine intelligence, 41(7):1641–1654, 2018

    2018

  46. [52]

    Are we ready for autonomous driving? the kitti vision benchmark suite

    Andreas Geiger, Philip Lenz, and Raquel Urtasun. Are we ready for autonomous driving? the kitti vision benchmark suite. In CVPR, 2012

    2012

  47. [53]

    The pascal visual object classes (voc) challenge

    Mark Everingham, Luc Van Gool, Christopher KI Williams, John Winn, and Andrew Zisserman. The pascal visual object classes (voc) challenge. International journal of computer vision, 88(2):303–338, 2010

    2010

  48. [54]

    The open images dataset v4: Unified image classification, object detection, and visual relationship detection at scale

    Alina Kuznetsova, Hassan Rom, Neil Alldrin, Jasper Ui- jlings, Ivan Krasin, Jordi Pont-Tuset, Shahab Kamali, Stefan Popov, Matteo Malloci, Tom Duerig, and Vittorio Ferrari. The open images dataset v4: Unified image classification, object detection, and visual relationship detect...

    2018 arXiv

  49. [55]

    Visual genome: Connecting language and vision using crowdsourced dense image annotations

    Ranjay Krishna, Yuke Zhu, Oliver Groth, Justin Johnson, Kenji Hata, Joshua Kravitz, Stephanie Chen, Yannis Kalan- tidis, Li-Jia Li, David A Shamma, et al. Visual genome: Connecting language and vision using crowdsourced dense image annotations. International Journal of Compute...

    2017

  50. [56]

    Imagenet: A large-scale hierarchical image database

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. Imagenet: A large-scale hierarchical image database. In CVPR, 2009

    2009

  51. [57]

    Wordnet: a lexical database for english

    George A Miller. Wordnet: a lexical database for english. Communications of the ACM, 38(11):39–41, 1995

    1995

  52. [61]

    Rep- met: Representative-based metric learning for classification and one-shot object detection

    Eli Schwartz, Leonid Karlinsky, Joseph Shtok, Sivan Harary, Mattias Marder, Sharathchandra Pankanti, Rogerio Feris, Abhishek Kumar, Raja Giries, and Alex M Bronstein. Rep- met: Representative-based metric learning for classification and one-shot object detection. In CVPR, 2019

    2019

  53. [62]

    Selective search for ob- ject recognition

    Jasper RR Uijlings, Koen EA Van De Sande, Theo Gev- ers, and Arnold WM Smeulders. Selective search for ob- ject recognition. International journal of computer vision , 104(2):154–171, 2013

    2013

  54. [63]

    Navaneeth Bodla, Bharat Singh, Rama Chellappa, and Larry S. Davis. Soft-nms improving object detection with one line of code. In ICCV, 2017

    2017

  55. [64]

    Arteta, V

    C. Arteta, V . Lempitsky, and A. Zisserman. Counting in the wild. In ECCV, 2016

    2016

  56. [65]

    The cityscapes dataset for semantic urban scene understanding

    Marius Cordts, Mohamed Omran, Sebastian Ramos, Timo Rehfeld, Markus Enzweiler, Rodrigo Benenson, Uwe Franke, Stefan Roth, and Bernt Schiele. The cityscapes dataset for semantic urban scene understanding. In CVPR, 2016

    2016

  57. [66]

    Domain adaptive faster r-cnn for object de- tection in the wild

    Yuhua Chen, Wen Li, Christos Sakaridis, Dengxin Dai, and Luc Van Gool. Domain adaptive faster r-cnn for object de- tection in the wild. In CVPR, 2018

    2018

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.