Pith. sign in

REVIEW 3 major objections 5 minor 51 references

SimLTD: Simple Supervised and Semi-Supervised Long-Tailed Object Detection

T0 review · 3 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read SimLTD claims that a simple three-stage head-to-tail transfer, optionally using unlabeled images via pseudo-labeling, achieves new record results on LVIS v1 without any extra image-level labels or pretrained vision-language models.

desk verdict A clean three-stage recipe that gets strong LVIS numbers without ImageNet/CLIP, but the semi-supervised headline overreaches and one key ablation is missing. read the letter →

arxiv 2412.20047 v3 pith:ZDPO7G7E submitted 2024-12-28 cs.CV cs.LG

Phi Vu Tran This is my paper · ORCID
classification cs.CVcs.LG
keywords long-tailedobjectdetectionsemi-supervisedhead-to-tailtransferpseudo-labelingLVISbenchmarkclassimbalancefew-shotlearningdataaugmentation
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

The paper argues that long-tailed object detection does not need a huge external labeled database. It proposes SimLTD, a three-stage recipe that pre-trains on head classes, transfers to tail classes, then fine-tunes on a balanced sample, optionally using unlabeled images through pseudo-labeling. On the LVIS v1 benchmark, this simple pipeline establishes new record results under both fully supervised and semi-supervised settings, beating methods that rely on vision-language models and millions of extra labeled images. The implication is that easily collected unlabeled images can substitute for costly human annotations in boosting rare-class accuracy.

What carries the argument

The central mechanism is the three-stage head-to-tail transfer with head-tail class fusion. Stage one trains a detector on the 866 head classes (with repeat-factor sampling and copy-paste). Stage two copies those weights and fine-tunes only the box classifier and regressor on the 337 tail classes, optionally with unlabeled images whose rare instances are augmented by pasting labeled tail objects into diverse scenes so the teacher sees enough tail examples to propose pseudo-labels. Stage three merges the two detectors' classifier and regressor weights via a simple index map (Algorithm 1) and fine-tunes on a 30-shot-per-class replay set to prevent catastrophic forgetting.

What would settle it

Train the semi-supervised pipeline identically but disable the rare-instance pasting in Step 2 (or paste head-class objects instead), and compare rare-class APr on LVIS v1; a non-material drop would show that the pasted-object pseudo-labels, not the multi-stage curriculum, are responsible for the reported gains.

Watch

Extended reading notes

Core claim

SimLTD's central discovery is that decoupling learning into head pre-training, tail transfer, and balanced fine-tuning, with pseudo-labels on unlabeled images, is enough to match or exceed far more complex systems. The paper shows empirically that stronger COCO-pretrained representations transfer better to LVIS rare classes, then builds a training curriculum that exploits this. Adding unlabeled images improves both head representation learning and tail adaptation, and the method scales across convolutional and transformer detectors. Removing the external labeled-image crutch does not hurt; in the fully supervised setting SimLTD already outperforms those label-hungry competitors on rare-class AP.

Load-bearing premise

The load-bearing premise is that pasting rare labeled instances into unlabeled images produces pseudo-labels that genuinely improve tail-class adaptation; if those pasted scenes are unrealistic or over-represent the tail categories, the reported transfer gains could be an artifact of the augmentation rather than a real improvement in learning rare-class features.

Editorial extensions

If this is right

  • Long-tailed detection can be advanced with unlabeled images alone, removing the requirement of building or licensing a large object-centric labeled corpus.
  • The same three-stage recipe transfers across detectors and backbones, from Faster R-CNN to DINO with Swin backbones, so practitioners can apply it to their own architectures.
  • Semi-supervised SimLTD exceeds previous methods that use ImageNet labels and a CLIP classifier on rare-class AP with ResNet backbones, and matches them with larger Swin backbones.
  • Supervised SimLTD, using no external data at all, already outperforms earlier multi-stage long-tailed detectors such as the three-step smooth-tail method using Deformable DETR.
  • The 30-shot fine-tuning sweet spot offers a practical rule for balancing head-class preservation and tail-class accuracy.

Reading between the lines

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

  • A natural extension not explored in the paper: the copy-paste of tail instances into unlabeled scenes is a cheap form of synthetic oversampling of rare classes, and similar mechanisms might improve other imbalanced recognition tasks like instance segmentation or open-vocabulary detection.
  • If the gains hold, the dependence of prior long-tailed methods on large pretrained vision-language models may be a workaround for a weak training curriculum rather than a necessary ingredient, suggesting that stronger curricula could further close the gap with fully supervised data.
  • One testable extension is to apply the same head-to-tail transfer to bespoke domains like aerial or medical imagery, where the paper argues no equivalent of ImageNet exists but large unlabeled corpora do.
  • The method's reliance on pseudo-label correctness means that better teacher models or calibrated confidence thresholds should directly improve tail transfer, a prediction that can be checked by ablating the teacher's accuracy.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes SimLTD, a three-stage pipeline for long-tailed object detection: (1) pre-training on head classes, (2) transfer learning on tail classes by fine-tuning only the classifier and regressor, and (3) fine-tuning on a sampled set Dk of k instances per class. The method optionally uses unlabeled images via pseudo-labeling and introduces a procedure that pastes rare training instances into unlabeled images to increase pseudo-labels for tail classes. Experiments on LVIS v1 report strong mAPbox and APr across Faster R-CNN, Deformable DETR, and DINO with ResNet/Swin backbones, including semi-supervised settings with COCO-unlabeled and Objects365-unlabeled images, plus an ancillary Objects365 evaluation. The paper claims new state-of-the-art results on LVIS v1 without external image-level labels.

Significance. If the results hold, SimLTD is significant because it demonstrates that a simple multi-stage head-to-tail transfer with unlabeled pseudo-labels can compete with or beat methods that rely on CLIP and roughly 1.5M labeled ImageNet images, which would broaden the applicability of long-tailed detection to domains without large object-centric databases. The paper is also valuable for its systematic comparisons across detectors and backbones, its APFixed comparison to CascadeMatch, and the Objects365 generalization check. However, the record claims currently rest on an unverified pasting component, point estimates without variance, and hyperparameters selected on the evaluation set, so the significance is conditional on completing the missing ablation and qualifying the claims.

major comments (3)
  1. [Section 3.3 and Section 4.5] Section 3.3 introduces rare-instance pasting and justifies it by saying that 'our ablation experiments in §4.5 reveal that it is surprisingly helpful,' but Section 4.5 contains no such ablation. Table 4 toggles RFS, Copy-Paste, and multi-stage training in the fully supervised setting, and Figure 8 toggles tail initialization and the number of shots k; neither isolates the paste/no-paste decision in the semi-supervised pipeline. Because the pasted objects are rare training instances with ground-truth boxes, the pseudo-label loss in Step 2 is effectively supervised by those exact exemplars repeated over many backgrounds, so the reported APr gain could reflect memorization of specific training instances rather than transfer from unlabeled data. Please add a paste/no-paste ablation in the semi-supervised setting and, ideally, report pseudo-label statistics such as the number of rare-class pseudo-boxes and their precision to show that the augmented images teach transferable features rather than unrealistic image statistics.
  2. [Table 2 and Abstract] The claim of 'new record results' on LVIS v1 is too strong as stated. In the semi-supervised Swin-L row, SimLTD MixPL with COCO-unlabeled reaches 51.5 mAPbox / 45.0 APr, while RichSem with ImageNet-21K and CLIP reaches 52.0 / 50.2, so this configuration is not a record. In the supervised Swin-L row, the margin over RichSem is only 0.1 mAPbox (49.8 vs 49.7). Since Section 4.1 states that Dk is sampled three times, please report the variance or standard deviation for the main results and either soften the 'new record' claim or restrict it to the configurations where the evidence actually supports it.
  3. [Section 4.5, Figure 8] The fine-tuning shot count k = 30 is selected by examining the 'knee in the curve' on the same LVIS v1 validation set used to report all final numbers in Table 2. This makes the reported metrics optimistically biased, which is particularly problematic for a paper whose central claim is a new record by small margins. Please report results on a held-out split, or at least use a validation split only for hyperparameter selection, or include a sensitivity analysis showing that the qualitative conclusions are stable across reasonable choices of k.
minor comments (5)
  1. [Section 4.3, Objects365] The phrase 'k = 30000 shots' is inconsistent with the per-class definition of k introduced in Section 3; with 365 classes this would imply a very large number of per-class samples. Please clarify whether k is per class or total, and how this value relates to the 30-shot choice on LVIS.
  2. [Figure 3 caption] The caption says 'The model is updated only during pre-training,' which contradicts Sections 3.3 and 3.4 where the classifier and regressor are updated in Steps 2 and 3. Please clarify that the backbone or encoder-decoder is frozen after Step 1, while the detector head is updated later.
  3. [Section 4.2] The statement that the semi-supervised methods 'inherit all hyper-parameters originally tuned on the COCO dataset' leaves the pseudo-label loss weight alpha from Equation (2) and the teacher-update schedule unspecified. Please provide these values or point to the exact configuration in the released code.
  4. [Abstract and Section 4.3] The phrase 'without extra image labels' should be qualified because the Swin backbones are pre-trained on ImageNet-22K, which uses image-level labels. The claim is understandable as 'without additional training-image labels beyond standard backbone pre-training,' but the current wording is easily misread.
  5. [Figure 8, left panel] The annotations '+25%' and '+19%' are not defined; please specify whether they are relative or absolute AP changes and over which baseline they are computed.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the three-stage training method is evaluated on held-out LVIS v1 validation, and the author's own SoftER Teacher is a non-load-bearing component, not the source of the headline results.

full rationale

SimLTD is an empirical method paper: its three-step recipe (head pre-training, tail transfer, fused fine-tuning) is evaluated against the held-out LVIS v1 validation set and compared with externally reported baselines, so no result is defined in terms of its own input or fitted value. The only self-citation is SoftER Teacher [35], used as one of three semi-supervised objectives; the headline semi-supervised numbers use MixPL, and SoftER Teacher is an independently published method, so this citation does not carry the central claim. The rare-instance pasting in §3.3 propagates existing ground-truth boxes into unlabeled images as a training augmentation; it is an input transformation, not a prediction derived from the model, and the SOTA claim is not a mathematical consequence of it. The k=30 choice is a validation-based hyperparameter selection reported with seeded averaging, not a fitted parameter renamed as a prediction. Although §3.3 references an ablation in §4.5 that does not explicitly isolate the pasting component, that is a verification/completeness concern, not circularity. No equation is self-referential, no uniqueness theorem is imported from the authors' prior work, and no known result is repackaged under new names. Therefore the derivation chain is self-contained and no circular step can be exhibited.

Assumptions & free parameters 3 free parameters · 3 assumptions · 0 invented entities

The framework has no new scientific entities, but it carries several hand-picked or borrowed hyperparameters and rests on two empirical premises. The k-shot count is fit to the target benchmark, which is the most consequential free parameter; the copy-paste into unlabeled images is a novel trick whose benefit is only shown empirically.

free parameters (3)
  • k (number of shots for fine-tuning Dk) = 30 on LVIS v1; 30000 on Objects365
    Chosen on the target validation set by analyzing the knee in the mAP/APr curves (Figure 8); it is a data-dependent hyperparameter that scales with dataset, so the reported benchmark numbers reflect this tuned value.
  • Head-tail split threshold M = 10 images for LVIS v1; 100 for Objects365
    Defines which classes are head versus tail; follows the LVIS protocol in prior work, but is a dataset-specific choice that shapes the training stages and the final results.
  • Pseudo-label loss weight alpha = Inherited from SoftER Teacher, MixTeacher, MixPL
    Controls the contribution of unlabeled data in Equation (2); the paper does not tune it, but the semi-supervised results depend on this borrowed hyperparameter.
assumptions (3)
  • domain assumption Stronger pre-trained representations (as measured by COCO AP) lead to better downstream rare-class detection on LVIS.
    Empirical premise established only by the small experiment in Section 3.1 (Figure 4); the whole head-to-tail transfer design rests on this monotonicity, which is shown for a few detectors without statistical testing.
  • ad hoc to paper Pasting rare instances from the labeled set into unlabeled images provides the teacher model with useful pseudo-labels that improve tail adaptation, rather than teaching the detector unrealistic image statistics.
    Introduced in Section 3.3 and supported only by the ablation in Section 4.5; no theoretical or measurement evidence that the augmented scenes preserve the target distribution.
  • domain assumption Student-teacher semi-supervised methods tuned on COCO (SoftER Teacher, MixTeacher, MixPL) transfer to LVIS without hyperparameter changes.
    The paper states all hyperparameters are inherited 'without changes' (Section 4.2), so the reported semi-supervised gains rest on this transferability assumption.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SimLTD: Simple Supervised and Semi-Supervised Long-Tailed Object Detection." pith.science (2026). https://pith.science/paper/ZDPO7G7E

@misc{pith2026241220047,
  author       = {Pith},
  title        = {Pith review of: SimLTD: Simple Supervised and Semi-Supervised Long-Tailed Object Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZDPO7G7E}},
  note         = {Machine review of arXiv:2412.20047}
}
read the original abstract

While modern visual recognition systems have made significant advancements, many continue to struggle with the open problem of learning from few exemplars. This paper focuses on the task of object detection in the setting where object classes follow a natural long-tailed distribution. Existing methods for long-tailed detection resort to external ImageNet labels to augment the low-shot training instances. However, such dependency on a large labeled database has limited utility in practical scenarios. We propose a versatile and scalable approach to leverage optional unlabeled images, which are easy to collect without the burden of human annotations. Our SimLTD framework is straightforward and intuitive, and consists of three simple steps: (1) pre-training on abundant head classes; (2) transfer learning on scarce tail classes; and (3) fine-tuning on a sampled set of both head and tail classes. Our approach can be viewed as an improved head-to-tail model transfer paradigm without the added complexities of meta-learning or knowledge distillation, as was required in past research. By harnessing supplementary unlabeled images, without extra image labels, SimLTD establishes new record results on the challenging LVIS v1 benchmark across both supervised and semi-supervised settings.

Figures

Figures reproduced from arXiv: 2412.20047 by the authors.

Figure 1
Figure 1. A survey comparing our SimLTD to the state of the art for [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The motivation to our approach. Left: We propose an improved head-to-tail model transfer framework for long-tailed detection by incorporating unlabeled images in both representation and transfer learning stages. Right: While it is possible to find more samples of LVIS instances from ImageNet, such an auxiliary database may not exist in another scenario like aerial imagery. Our framework does not depend on using extr… view at source ↗
Figure 3
Figure 3. Overview of SimLTD. Step 1 pre-trains the model and detector on head classes with unlabeled images. Step 2 transfers the head representations to tail classes. Step 3 fine-tunes the detector on a sample of head and tail exemplar replay. The “model” is an abstract block of encoder-decoder transformations based on either a convolutional or transformer network. The model is updated only during pre-training. and fine-tun… view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Transfer learning from COCO representations (solid bars) [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 5
Figure 5. Figure 5: Left: Augmenting unlabeled images with randomly pasted rare instances helps promote pseudo-labeling for effective semi￾supervised learning. Middle: The pasted objects (cyan boxes) often come from contrasting environments to create complex fake scenes for the student to…
Figure 6
Figure 6. Figure 6: SimLTD detections on LVIS v1 validation images. We highlight visualizations containing truly rare [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Evaluation on Objects365 binned by the count of training [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]
Figure 8
Figure 8. Figure 8: Ablation experiments assessing the impact on AP from [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

51 extracted references · 44 canonical work pages

  1. [1]

    YOLOv4: Optimal Speed and Accuracy of Object Detection

    Alexey Bochkovskiy, Chien-Yao Wang, and Hong-Yuan Mark Liao. YOLOv4: Optimal Speed and Accuracy of Object Detection. https://arxiv.org/abs/2004.10934, 2020. 3, 5

    arXiv 2004

  2. [2]

    Cascade R-CNN: Delv- ing into High Quality Object Detection

    Zhaowei Cai and Nuno Vasconcelos. Cascade R-CNN: Delv- ing into High Quality Object Detection. In CVPR, 2018. 7

    work page 2018

  3. [3]

    End-to- End Object Detection with Transformers

    Nicolas Carion, Francisco Massa, Gabriel Synnaeve, Nicolas Usunier, Alexander Kirillov, and Sergey Zagoruyko. End-to- End Object Detection with Transformers. In ECCV, 2020. 4

    work page 2020

  4. [4]

    Castro, Manuel J

    Francisco M. Castro, Manuel J. Marín-Jiménez, Nicolás Guil, Cordelia Schmid, and Karteek Alahari. End-to-End Incre- mental Learning. In ECCV, 2018. 1

    work page 2018

  5. [5]

    MMDetection: Open MMLab Detection Toolbox and Benchmark

    Kai Chen, Jiaqi Wang, Jiangmiao Pang, Yuhang Cao, Yu Xiong, Xiaoxiao Li, et al. MMDetection: Open MMLab Detection Toolbox and Benchmark. https://arxiv.org/abs/1906.07155, 2019. 6

    arXiv 1906

  6. [6]

    Dif- fusionDet: Diffusion Model for Object Detection

    Shoufa Chen, Peize Sun, Yibing Song, and Ping Luo. Dif- fusionDet: Diffusion Model for Object Detection. In ICCV,

    work page

  7. [7]

    Mixed Pseudo Labels for Semi-Supervised Object Detection

    Zeming Chen, Wenwei Zhang, Xinjiang Wang, Kai Chen, and Zhi Wang. Mixed Pseudo Labels for Semi-Supervised Object Detection. https://arxiv.org/abs/2312.07006, 2023. 4, 6

    arXiv 2023

  8. [8]

    Evaluating Large- V ocabulary Object Detectors: The Devil is in the Details

    Achal Dave, Piotr Dollár, Deva Ramanan, Alexan- der Kirillov, and Ross Girshick. Evaluating Large- V ocabulary Object Detectors: The Devil is in the Details. https://arxiv.org/abs/2102.01066, 2021. 7

    arXiv 2021

Show all 51 references
  1. [9]

    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. 2

    2009

  2. [10]

    Terrance DeVries and Graham W. Taylor. Improved Reg- ularization of Convolutional Neural Networks with Cutout. https://arxiv.org/abs/1708.04552, 2017. 5

    2017 arXiv

  3. [11]

    Object Detection in Aerial Images: A Large-Scale Benchmark and Challenges

    Jian Ding, Nan Xue, Gui-Song Xia, Xiang Bai, Wen Yang, Micheal Ying Yang, Serge Belongie, Jiebo Luo, Mihai Datcu, Marcello Pelillo, and Liangpei Zhang. Object Detection in Aerial Images: A Large-Scale Benchmark and Challenges. IEEE TPAMI, 2021. 2

    2021

  4. [12]

    Boosting Long-Tailed Object Detection via Step-Wise Learning on Smooth-Tail Data

    Na Dong, Yongqiang Zhang, Mingli Ding, and Gim Hee Lee. Boosting Long-Tailed Object Detection via Step-Wise Learning on Smooth-Tail Data. In ICCV, 2023. 1, 2, 4, 6, 8

    2023

  5. [13]

    Williams, John Winn, and Andrew Zisserman

    Mark Everingham, Luc Van Gool, Christopher K.I. Williams, John Winn, and Andrew Zisserman. The PASCAL Visual Object Classes (VOC) Challenge. IJCV, 2010. 1

    2010

  6. [14]

    Cubuk, Quoc V

    Golnaz Ghiasi, Yin Cui, Aravind Srinivas, Rui Qian, Tsung- Yi Lin, Ekin D. Cubuk, Quoc V . Le, and Barret Zoph. Simple Copy-Paste is a Strong Data Augmentation Method for In- stance Segmentation. In CVPR, 2021. 4

    2021

  7. [15]

    LVIS: A Dataset for Large V ocabulary Instance Segmentation

    Agrim Gupta, Piotr Dollár, and Ross Girshick. LVIS: A Dataset for Large V ocabulary Instance Segmentation. In CVPR, 2019. 1, 4

    2019

  8. [16]

    Deep Residual Learning for Image Recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep Residual Learning for Image Recognition. In CVPR,

  9. [17]

    Mask R-CNN

    Kaiming He, Georgia Gkioxari, Piotr Dollár, and Ross Gir- shick. Mask R-CNN. In ICCV, 2017. 4

    2017

  10. [18]

    The Devil is in the Tails: Fine-Grained Classification in the Wild

    Grant Van Horn and Pietro Perona. The Devil is in the Tails: Fine-Grained Classification in the Wild. https://arxiv.org/abs/1709.01450, 2017. 2

    2017 arXiv

  11. [19]

    Learning to Segment the Tail

    Xinting Hu, Yi Jiang, Kaihua Tang, Jingyuan Chen, Chunyan Miao, and Hanwang Zhang. Learning to Segment the Tail. In CVPR, 2020. 1, 2, 3, 4

    2020

  12. [20]

    Adaptive Hierarchical Representation Learning for Long-Tailed Object Detection

    Banghuai Li. Adaptive Hierarchical Representation Learning for Long-Tailed Object Detection. In CVPR, 2022. 6

    2022

  13. [21]

    Equalized Focal Loss for Dense Long-Tailed Object Detection

    Bo Li, Yongqiang Yao, Jingru Tan, Gang Zhang, Fengwei Yu, Jianwei Lu, and Ye Luo. Equalized Focal Loss for Dense Long-Tailed Object Detection. In CVPR, 2022. 6

    2022

  14. [22]

    Lawrence Zitnick

    Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Dollár, and C. Lawrence Zitnick. Microsoft COCO: Common Objects in Context. In ECCV, 2014. 1

    2014

  15. [23]

    Feature Pyramid Networks for Object Detection

    Tsung-Yi Lin, Piotr Dollár, Ross Girshick, Kaiming He, Bharath Hariharan, and Serge Belongie. Feature Pyramid Networks for Object Detection. In CVPR, 2017. 6

    2017

  16. [24]

    Focal Loss for Dense Object Detection

    Tsung-Yi Lin, Priya Goyal, Ross Girshick, Kaiming He, and Piotr Dollár. Focal Loss for Dense Object Detection. InICCV,

  17. [25]

    MixTeacher: Mining Promising La- bels with Mixed Scale Teacher for Semi-Supervised Object Detection

    Liang Liu, Boshen Zhang, Jiangning Zhang, Wuhao Zhang, Zhenye Gan, Guanzhong Tian, Wenbing Zhu, Yabiao Wang, and Chengjie Wang. MixTeacher: Mining Promising La- bels with Mixed Scale Teacher for Semi-Supervised Object Detection. In CVPR, 2023. 4, 6

    2023

  18. [26]

    Swin Transformer: Hierarchical Vision Transformer using Shifted Windows

    Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, and Baining Guo. Swin Transformer: Hierarchical Vision Transformer using Shifted Windows. In ICCV, 2021. 6

    2021

  19. [27]

    Learning from Rich Semantics and Coarse Locations for Long-Tailed Object Detection

    Lingchen Meng, Xiyang Dai, Jianwei Yang, Dongdong Chen, Yinpeng Chen, Mengchen Liu, Yi-Ling Chen, Zuxuan Wu, Lu Yuan, and Yu-Gang Jiang. Learning from Rich Semantics and Coarse Locations for Long-Tailed Object Detection. In NeurIPS, 2023. 2, 3, 6, 8

    2023

  20. [28]

    On Model Calibration for Long-Tailed Object Detection and Instance Segmentation

    Tai-Yu Pan, Cheng Zhang, Yandong Li, Hexiang Hu, Dong Xuan, Soravit Changpinyo, Boqing Gong, and Wei-Lun Chao. On Model Calibration for Long-Tailed Object Detection and Instance Segmentation. In NeurIPS, 2021. 6, 7

    2021

  21. [29]

    PyTorch: An Imperative Style, High- Performance Deep Learning Library

    Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, et al. PyTorch: An Imperative Style, High- Performance Deep Learning Library. In NeurIPS. Curran Associates, Inc., 2019. 6

    2019

  22. [30]

    Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger, and Ilya Sutskever

    Alec Radford, Jong Wook Kim, Chris Hallacy, A. Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger, and Ilya Sutskever. Learning Transferable Visual Models From Natural Language Supervision. In ICML, 2021. 3 9

    2021

  23. [31]

    Sylvestre-Alvise Rebuffi, Alexander Kolesnikov, Georg Sperl, and Christoph H. Lampert. iCaRL: Incremental Classifier and Representation Learning. In CVPR, 2017. 1

    2017

  24. [32]

    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. 4

    2015

  25. [33]

    Objects365: A Large-Scale, High-Quality Dataset for Object Detection

    Shuai Shao, Zeming Li, Tianyuan Zhang, Chao Peng, Gang Yu, Xiangyu Zhang, Jing Li, and Jian Sun. Objects365: A Large-Scale, High-Quality Dataset for Object Detection. In ICCV, 2019. 7

    2019

  26. [34]

    Mean Teachers are Better Role Models: Weight-Averaged Consistency Targets Improve Semi-Supervised Deep Learning Results

    Antti Tarvainen and Harri Valpola. Mean Teachers are Better Role Models: Weight-Averaged Consistency Targets Improve Semi-Supervised Deep Learning Results. In NeurIPS, 2017. 5

    2017

  27. [35]

    LEDetection: A Simple Framework for Semi- Supervised Few-Shot Object Detection

    Phi Vu Tran. LEDetection: A Simple Framework for Semi- Supervised Few-Shot Object Detection. In AISTATS, 2024. 3, 4, 6

    2024

  28. [36]

    Seesaw Loss for Long- Tailed Instance Segmentation

    Jiaqi Wang, Wenwei Zhang, Yuhang Zang, Yuhang Cao, Jiangmiao Pang, Tao Gong, Kai Chen, Ziwei Liu, Chen Change Loy, and Dahua Lin. Seesaw Loss for Long- Tailed Instance Segmentation. In CVPR, 2021. 6

    2021

  29. [37]

    The Devil is in Classification: A Simple Framework for Long-Tail Object Detection and Instance Segmentation

    Tao Wang, Yu Li, Bingyi Kang, Junnan Li, Junhao Liew, Sheng Tang, Steven Hoi, and Jiashi Feng. The Devil is in Classification: A Simple Framework for Long-Tail Object Detection and Instance Segmentation. In ECCV, 2020. 4

    2020

  30. [38]

    Huang, Trevor Darrell, Joseph E

    Xin Wang, Thomas E. Huang, Trevor Darrell, Joseph E. Gon- zalez, and Fisher Yu. Frustratingly Simple Few-Shot Object Detection. In ICML, 2020. 3

    2020

  31. [39]

    Learn- ing to Model the Tail

    Yu-Xiong Wang, Deva Ramanan, and Martial Hebert. Learn- ing to Model the Tail. In NeurIPS, 2017. 2, 3, 4

    2017

  32. [40]

    Large Scale In- cremental Learning

    Yue Wu, Yinpeng Chen, Lijuan Wang, Yuancheng Ye, Zicheng Liu, Yandong Guo, and Yun Fu. Large Scale In- cremental Learning. In CVPR, 2019. 1

    2019

  33. [41]

    End-to- End Semi-Supervised Object Detection with Soft Teacher

    Mengde Xu, Zheng Zhang, Han Hu, Jianfeng Wang, Lijuan Wang, Fangyun Wei, Xiang Bai, and Zicheng Liu. End-to- End Semi-Supervised Object Detection with Soft Teacher. In ICCV, 2021. 4

    2021

  34. [42]

    Instance-Aware Repeat Factor Sampling for Long-Tailed Ob- ject Detection

    Burhaneddin Yaman, Tanvir Mahmud, and Chun-Hao Liu. Instance-Aware Repeat Factor Sampling for Long-Tailed Ob- ject Detection. https://arxiv.org/abs/2305.08069, 2023. 4

    2023 arXiv

  35. [43]

    FASA: Feature Augmentation and Sampling Adaptation for Long- Tailed Instance Segmentation

    Yuhang Zang, Chen Huang, and Chen Change Loy. FASA: Feature Augmentation and Sampling Adaptation for Long- Tailed Instance Segmentation. In ICCV, 2021. 7

    2021

  36. [44]

    Semi-Supervised and Long-Tailed Object Detection with CascadeMatch

    Yuhang Zang, Kaiyang Zhou, Chen Huang, and Chen Change Loy. Semi-Supervised and Long-Tailed Object Detection with CascadeMatch. IJCV, 2023. 3, 7

    2023

  37. [45]

    MosaicOS: A Simple and Effective Use of Object-Centric Images for Long-Tailed Object Detection

    Cheng Zhang, Tai-Yu Pan, Yandong Li, Hexiang Hu, Dong Xuan, Soravit Changpinyo, Boqing Gong, and Wei-Lun Chao. MosaicOS: A Simple and Effective Use of Object-Centric Images for Long-Tailed Object Detection. In ICCV, 2021. 2, 3, 6

    2021

  38. [46]

    Dauphin, and David Lopez-Paz

    Hongyi Zhang, Moustapha Cisse, Yann N. Dauphin, and David Lopez-Paz. mixup: Beyond Empirical Risk Minimiza- tion. In ICLR, 2018. 5

    2018

  39. [47]

    Ni, and Heung-Yeung Shum

    Hao Zhang, Feng Li, Shilong Liu, Lei Zhang, Hang Su, Jun Zhu, Lionel M. Ni, and Heung-Yeung Shum. DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Ob- ject Detection. In ICLR, 2023. 4

    2023

  40. [48]

    Random Erasing Data Augmentation

    Zhun Zhong, Liang Zheng, Guoliang Kang, Shaozi Li, and Yi Yang. Random Erasing Data Augmentation. In AAAI, 2020. 5

    2020

  41. [49]

    Detecting Twenty-Thousand Classes Using Image-Level Supervision

    Xingyi Zhou, Rohit Girdhar, Armand Joulin, Philipp Krähen- bühl, and Ishan Misra. Detecting Twenty-Thousand Classes Using Image-Level Supervision. In ECCV, 2022. 2, 3, 6

    2022

  42. [50]

    Deformable DETR: Deformable Transformers for End-to-End Object Detection

    Xizhou Zhu, Weijie Su, Lewei Lu, Bin Li, Xiaogang Wang, and Jifeng Dai. Deformable DETR: Deformable Transformers for End-to-End Object Detection. In ICLR, 2021. 4

    2021

  43. [51]

    The Psycho-Biology of Language: An Introduction to Dynamic Philology

    George Kingsley Zipf. The Psycho-Biology of Language: An Introduction to Dynamic Philology. Routledge, 2013. 3 10

    2013

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.