Pith. sign in

REVIEW 4 major objections 7 minor 46 references

Pseudo-Label Quality Decoupling and Correction for Semi-Supervised Instance Segmentation

T0 review · 4 major / 7 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read This paper claims that decoupling class-quality and mask-quality thresholds for pseudo-label filtering, adding CLIP-based category correction, and reweighting mask loss by pixel uncertainty together set new state-of-the-art results in…

desk verdict A credible SSIS framework whose CLIP-based category correction is the real novelty, but the SOTA deltas are provisional given no code, no seeds, and a known small-object bias in DDTF. read the letter →

arxiv 2505.11075 v1 pith:BECAQSPF submitted 2025-05-16 cs.CV

Jianghang Lin , Yilin Lu , Yunhang Shen , Chaoyang Zhu , Shengchuan Zhang , Liujuan Cao , Rongrong Ji This is my paper · ORCID
classification cs.CV
keywords semi-supervisedinstancesegmentationpseudo-labelfilteringdual-thresholddecouplingclassqualitymaskCLIPcategorycorrectionpixeluncertaintyweightingteacher-studentlearning
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 targets semi-supervised instance segmentation, where a model is trained on a few labeled images together with many unlabeled ones to produce object masks and their categories. Its central claim is that pseudo-label filtering should not use a single threshold on the product of class confidence and mask confidence, because that product forces a trade-off between classification quality and mask quality; instead the two qualities should be gated independently. The proposed framework couples this decoupled dual-threshold filter with a CLIP-based category-correction module whose influence fades as the teacher improves, and with a per-pixel uncertainty weighting that down-weights ambiguous mask predictions in the loss. The authors report state-of-the-art results on COCO and Cityscapes at every tested label ratio, with the largest gains under extreme scarcity (+11.6 mAP at 1% COCO labels and +15.5 mAP at 5% Cityscapes labels). If these results hold, the practical bottleneck in semi-supervised instance segmentation is not the amount of pseudo-label data but its quality control.

What carries the argument

The central object is the decoupled dual-threshold filter: rather than one threshold on the coupled score $s_k=c_k \cdot m_k$, the teacher's pseudo-labels are accepted only when class confidence $c_k \ge c_t$ and mask confidence $m_k \ge m_t$ independently, with $m_t=0.9$ and $c_t=0.85$. The mask confidence $m_k$ is the mean sigmoid foreground probability over pixels predicted as foreground, defined in Eq. 4. Two mechanisms complete the pipeline: Dynamic Instance Category Correction (DICC) blends the teacher's class distribution with CLIP's zero-shot patch-level class distribution using a weight that decays from 0.5 to 0, and Pixel-level Mask Uncertainty-Aware (PMUA) weights each pixel's mask loss by $1-u_i^k$, where $u_i^k = 1 - 2|\sigma(q_i^k)-0.5|$ captures how near the teacher's prediction is to the decision boundary. DDTF is the entry gate that determines which instances enter the training pool, DICC repairs their category labels, and PMUA suppresses the noisy pixel-level supervision inside the kept masks.

What would settle it

Measure the Spearman rank correlation between the teacher's mask-quality score $m_k$ (Eq. 4) and the IoU of the predicted mask against the ground-truth mask, computed on a held-out set of unlabeled COCO images with oracle masks, and repeat for small, medium, and large objects; if for small objects the correlation is near zero or negative, DDTF's fixed $m_t=0.9$ would filter out precisely the best small-instance pseudo-labels, and the claimed small-object gains would fail to reproduce.

Watch

Extended reading notes

Core claim

At the heart of the paper is the observation that an instance pseudo-label carries two independent qualities - a class quality (how confident the model is about the category) and a mask quality (how confident it is about the pixel-level foreground) - and that their product, the standard instance score, does not reliably rank pseudo-labels by true mask IoU. The authors argue that filtering with a single threshold on this product sets up a competition between the two qualities, so they propose the Pseudo-Label Quality Decoupling and Correction (PL-DC) framework. Its Decoupled Dual-Threshold Filtering keeps a pseudo-label only when the class confidence meets $c_t=0.85$ and the mask confidence meets $m_t=0.9$ independently. Its Dynamic Instance Category Correction module feeds the masked image patch to CLIP, fuses CLIP's zero-shot category distribution with the teacher's predicted distribution under a cosine-decayed weight that starts at 0.5 and falls to 0, and takes the argmax as the corrected category. Its Pixel-level Mask Uncertainty-Aware loss reweights each pixel's mask cross-entropy by $1-u_i^k$, where $u_i^k$ is highest when teacher confidence is near 0.5. The paper claims that this three-part combination produces new state-of-the-art results in semi-supervised instance segmentation on COCO and Cityscapes, exceeding the previous best method at all evaluated labeled-data ratios.

Load-bearing premise

The load-bearing premise is that a teacher model's own confidence scores - especially the mask-quality score $m_k$, the mean foreground sigmoid probability - are honest estimates of pseudo-label quality on unlabeled data; if that confidence-to-quality link breaks, as the paper concedes it already does for small objects, the decoupled thresholds either admit noisy masks or discard good ones, and the reported gains are not guaranteed.

Editorial extensions

If this is right

  • If the results are right, the standard practice of a single confidence-product threshold in semi-supervised instance segmentation is suboptimal, and independent class and mask gates should be the default.
  • Extreme label scarcity becomes less forbidding: the reported +11.6 mAP at 1% COCO and +15.5 mAP at 5% Cityscapes imply that well-filtered pseudo-labels can nearly close the gap to much larger labeled sets.
  • The compatibility experiments indicate the category-correction and uncertainty modules are drop-in additions that improve other teacher-student SSIS frameworks as well, not just the paper's own combination.
  • Because all three modules act only during training, the inference-time cost of the model is unchanged; any gains come without extra test-time computation.

Reading between the lines

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

  • A natural extension the authors do not test: applying the same decoupling logic to the mask-quality score by learning a per-size or per-instance mask threshold rather than a fixed $m_t=0.9$, since their own ablation shows small objects are hurt by the fixed mask gate.
  • The same three-step treatment (separate gates, external-class correction, per-pixel confidence weighting) could transfer to other dense prediction tasks built on teacher-student pipelines, such as semi-supervised panoptic segmentation or video instance segmentation.
  • The DICC module's cosine decay assumes the teacher's class distribution monotonically improves; a testable consequence is that the optimal decay schedule should depend on per-category teacher accuracy, and a category-aware schedule might extend the gains.
  • The paper's evidence that the coupled score does not track IoU suggests a broader diagnostic: in any semi-supervised dense-prediction setup, one should verify that the confidence score used to filter pseudo-labels actually correlates with the ground-truth quality metric, not just with the teacher's self-assessment.
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

4 major / 7 minor

Summary. The manuscript proposes PL-DC, a framework for semi-supervised instance segmentation built on a Mask2Former teacher-student setup with EMA. Three contributions are introduced: (i) Decoupled Dual-Threshold Filtering (DDTF), which filters instance pseudo-labels using separate thresholds on class confidence c_k and mask quality m_k instead of the coupled product; (ii) Dynamic Instance Category Correction (DICC), which uses CLIP to correct pseudo-label categories with a cosine schedule that anneals the CLIP weight from 0.5 to 0; and (iii) Pixel-level Mask Uncertainty-Aware loss (PMUA), which reweights the pixel-wise mask BCE by per-pixel uncertainty u_i = 1 - 2|sigma(q_i) - 0.5|. Experiments on COCO (1%, 2%, 5%, 10%, 100% labeled) and Cityscapes (5%, 10%, 20%, 30%) report consistent improvements over GuidedDistillation and other SSIS methods, including headline gains of +11.6 mAP at 1% COCO and +15.5 mAP at 5% Cityscapes relative to supervised Mask2Former. The paper also includes module ablations, hyperparameter sweeps, compatibility experiments with prior SSIS methods, and a short-schedule convergence analysis.

Significance. If the reported gains are reproducible, this is a meaningful contribution to SSIS: it identifies a real limitation of single-score pseudo-label filtering, introduces CLIP-based category correction to SSIS, and shows through the compatibility experiments in Table 5 that the proposed modules are not specific to one detector family. The framework is not circular: the category correction uses an external pretrained model, and the Appendix C computation is a gradient calculation rather than a derivation of the SOTA numbers. The main weakness is experimental validation: all ablations and threshold choices are made on a 10-epoch, 1%-COCO schedule, the headline numbers are single-run without error bars or released code, and the paper itself concedes that the mask-quality threshold mis-filters small objects. These gaps currently prevent the SOTA claim from being fully established.

major comments (4)
  1. [§4.3, Table 3; Appendix E] The central DDTF claim is weakened by the paper's own small-object results. In Table 3, replacing DDTF with a coupled threshold of 0.765 raises AP_s by +0.3 mAP while lowering mAP by only 0.5, and the text in §4.3 and Appendix E explicitly states that the fixed mask-quality threshold 'adversely affects small object quality' and 'warrants further investigation.' Since Eq. (4) defines m_k as the mean sigmoid probability over confident foreground pixels and no size-dependent calibration is provided, the decoupling benefit is not established for small-object-heavy regimes. The authors should either introduce a size-adaptive mask-quality criterion or demonstrate that the AP_s regression disappears at the full training schedule.
  2. [§4.1 and §4.3] All module ablations and hyperparameter sweeps (Tables 3 and 4) are run for 73K iterations (10 epochs) on 1% COCO, while the headline results use 360K iterations on COCO and 180K on Cityscapes. The convergence curves in Figure II are also limited to 10 epochs, and Appendix E states that the effect of removing DDTF on small objects reverses during the course of training. Therefore the contribution of each module at the final operating point is unknown, and the attribution of the final SOTA gains to the three modules is not directly supported. The ablations should be repeated at the full schedule or at a substantially longer schedule to validate the modules' contributions.
  3. [§4.2, Tables 1-2] All reported numbers are single-run, with no error bars, seeds, or statistical significance tests. Several ablated differences in Table 4 are on the order of 0.4-0.8 mAP (e.g., c_t = 0.8 vs. 0.85, and alpha = 0.999 vs. 0.9996), which is within a plausible range of run-to-run variation for a short training with batch size 4; the same concern applies to the module differences in Table 3. Reporting variance over at least three seeds, or releasing code and logs so that the headline numbers can be reproduced, is necessary to support the claim of 'new state-of-the-art results.'
  4. [§4.3, Table 4] The thresholds m_t = 0.9 and c_t = 0.85 are selected by tuning on the same 1%-COCO, 10-epoch protocol that is used for the headline ablations, and are then applied to all COCO ratios, all Cityscapes ratios, and the full 360K/180K schedules without held-out validation. Table 4(a) shows that the selected thresholds are not robust in their neighborhood: both lower combinations (0.8/0.7) and the higher combination (0.9/0.9) produce notably worse mAP. The paper should either validate the thresholds on a separate split or schedule, or provide sensitivity analyses at the final training schedule, to rule out threshold-selection effects on the reported gains.
minor comments (7)
  1. [§4.3 and Table 1] The heading 'Abalation Study' should be 'Ablation Study', and 'Superised' in Table 1 is a typo for 'Supervised.'
  2. [§3.1, Eq. (4)] When no pixel satisfies sigma(q_i^k) > 0.5, the denominator in the mask-quality score m_k is zero; the authors should specify a default value for degenerate mask predictions.
  3. [Figure 2] The claim that the coupled instance score is not positively correlated with IoU is supported only by a qualitative scatter plot; please include Pearson or Spearman correlation coefficients for panels (a), (b), and (c).
  4. [Appendix C] The appendix is labeled a 'validity proof' but contains only a chain-rule computation of the gradient; since the derivation assumes the linear model t = theta x, which is not the actual Mask2Former output, the authors should present this as a simplified local analysis rather than a proof of robustness.
  5. [§4.4, Table 5] Each row of Table 5 (a-h) is described in a single sentence; adding a column that states the exact modification, and the component it replaces, would make the compatibility claims less ambiguous.
  6. [§3.2, Eq. (7)] The text says the CLIP weight w decays from 0.5 to 0, and Eq. (7) uses a cosine schedule; please confirm the schedule's exact range and state whether 'it_max' refers to the full 360K iterations or to the current training stage.
  7. [Appendix D] The CLIP precision/recall analysis uses ground-truth masks from COCO val; please state whether the same text template is used for all 80 categories and whether prompt ensembling is applied.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the reported gains are external benchmark results and no claimed prediction reduces to a fitted input or self-citation.

full rationale

The paper's load-bearing claims are empirical: DDTF, DICC, and PMUA are modules whose value is established by ablations and comparisons on COCO and Cityscapes against external baselines. Eq. 4 defines m_k as a mean foreground sigmoid probability, a confidence statistic; it is not defined in terms of the target mAP, and the paper explicitly concedes (Sec. 4.3 and Appendix E) that this fixed mask-quality threshold can hurt small objects, so the module's benefit is not guaranteed by construction. The Appendix C 'proof' is a direct chain-rule computation showing that the (1-u) factor in Eq. 11 dampens gradients when u is large; this is a property of the loss as defined, not an independent prediction, and no empirical conclusion is derived solely from it. Thresholds m_t and c_t are hyperparameters selected on the validation set (Table 4), which is standard tuning rather than a fitted parameter renamed as a prediction. Self-citations, notably PAIS [23] with overlapping authorship, serve as a baseline and are explicitly differentiated from DDTF; no load-bearing argument reduces to that citation. The central claim (SOTA mAP) is therefore externally falsifiable and not equivalent to an input by construction.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claim rests on a standard teacher-student self-training loop, a confidence-based filter, CLIP as an external oracle, and an uncertainty reweighting schedule. The first three are domain assumptions, the last is an ad hoc construction, and the thresholds and weights are tuned on validation data.

free parameters (5)
  • mask quality threshold m_t = 0.9
    Selected from Table 4(a) on COCO 1% validation; controls how many mask pseudo-labels survive filtering.
  • class quality threshold c_t = 0.85
    Selected from Table 4(a); controls category pseudo-label filtering.
  • EMA rate alpha = 0.9996
    Selected from Table 4(b); the central teacher-student update depends on this value.
  • unsupervised loss weight lambda = 1.0
    Selected from Table 4(c); performance collapses at lambda values of 4 and 8.
  • CLIP fusion schedule w = w = 0.25(cos(it_cur/it_max * pi) + 1)
    Ad hoc schedule decaying CLIP weight from 0.5 to 0; the 0.25 constant is not derived.
assumptions (4)
  • domain assumption Teacher-student EMA self-training improves semi-supervised instance segmentation in this setting.
    The framework's core training loop; inherited from UbTeacher-style methods and not re-derived here.
  • domain assumption The teacher's mask quality m_k (Eq. 4) is positively correlated with true mask IoU on unlabeled data.
    Required for DDTF to filter well; the authors note it is unreliable for small objects in Sec. 4.3 and Appendix E.
  • domain assumption CLIP's image-text alignment transfers to object patches and COCO category names well enough to correct pseudo-labels.
    DICC uses CLIP on masked patches; Appendix D shows per-category recall as low as 1.1% for knife.
  • ad hoc to paper Pixel uncertainty u_i = 1 - 2|sigma(q_i) - 0.5| is a useful proxy for mask-label noise.
    Defined in Eq. 10; the Appendix C 'proof' only shows gradient scaling, not that this weighting improves optimization.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Pseudo-Label Quality Decoupling and Correction for Semi-Supervised Instance Segmentation." pith.science (2026). https://pith.science/paper/BECAQSPF

@misc{pith2026250511075,
  author       = {Pith},
  title        = {Pith review of: Pseudo-Label Quality Decoupling and Correction for Semi-Supervised Instance Segmentation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BECAQSPF}},
  note         = {Machine review of arXiv:2505.11075}
}
read the original abstract

Semi-Supervised Instance Segmentation (SSIS) involves classifying and grouping image pixels into distinct object instances using limited labeled data. This learning paradigm usually faces a significant challenge of unstable performance caused by noisy pseudo-labels of instance categories and pixel masks. We find that the prevalent practice of filtering instance pseudo-labels assessing both class and mask quality with a single score threshold, frequently leads to compromises in the trade-off between the qualities of class and mask labels. In this paper, we introduce a novel Pseudo-Label Quality Decoupling and Correction (PL-DC) framework for SSIS to tackle the above challenges. Firstly, at the instance level, a decoupled dual-threshold filtering mechanism is designed to decouple class and mask quality estimations for instance-level pseudo-labels, thereby independently controlling pixel classifying and grouping qualities. Secondly, at the category level, we introduce a dynamic instance category correction module to dynamically correct the pseudo-labels of instance categories, effectively alleviating category confusion. Lastly, we introduce a pixel-level mask uncertainty-aware mechanism at the pixel level to re-weight the mask loss for different pixels, thereby reducing the impact of noise introduced by pixel-level mask pseudo-labels. Extensive experiments on the COCO and Cityscapes datasets demonstrate that the proposed PL-DC achieves significant performance improvements, setting new state-of-the-art results for SSIS. Notably, our PL-DC shows substantial gains even with minimal labeled data, achieving an improvement of +11.6 mAP with just 1% COCO labeled data and +15.5 mAP with 5% Cityscapes labeled data. The code will be public.

Figures

Figures reproduced from arXiv: 2505.11075 by the authors.

Figure 1
Figure 1. The proposed PL-DC outperforms the previous state [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The relationship between predicted instance scores and the IoU of predicted versus ground-truth instance masks. (a) Predicted [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Confusion matrix of the model trained on 1% COCO. For clarity, we visualize only the 29 most confused object cate￾gories and 1 background category. visual-language alignment model, CLIP, which has been pre-trained on large-scale image-text pairs. This module dy￾namically corrects the probability distribution of category pseudo-labels, effectively mitigating category confusion. Specifically, CLIP processes image patc… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Framework of our proposed pseudo-label quality decoupling and correction (PL-DC) for semi-supervised instance seg￾mentation. PL-DC includes two segmentation models, both Mask2Former [11], with identical configurations, namely Teacher and Student. The Teacher model gene…
Figure 5
Figure 5. Figure 5: Segmentation Analysis. We randomly sampled 1k images from the COCO train2017 dataset to analysis the segmentation results. They are categorized into 5 types: correct segmentation (Cor), poor localization (Loc), confusion with similar objects (Sim), confusion with objec…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

46 extracted references · 33 canonical work pages

  1. [1]

    Learn- ing with pseudo-ensembles.Advances in neural information processing systems, 27, 2014

    Philip Bachman, Ouais Alsharif, and Doina Precup. Learn- ing with pseudo-ensembles.Advances in neural information processing systems, 27, 2014. 3

    work page 2014

  2. [2]

    Deep watershed transform for instance segmentation

    Min Bai and Raquel Urtasun. Deep watershed transform for instance segmentation. InProceedings of the IEEE con- ference on computer vision and pattern recognition, pages 5221–5229, 2017. 3

    work page 2017

  3. [3]

    Guided distillation for semi-supervised instance segmentation

    Tariq Berrada, Camille Couprie, Karteek Alahari, and Jakob Verbeek. Guided distillation for semi-supervised instance segmentation. InProceedings of the IEEE/CVF Winter Con- ference on Applications of Computer Vision (WACV), pages 475–483, 2024. 1, 3, 4, 6, 7, 8

    work page 2024

  4. [4]

    Semi- supervised learning with distribution alignment and augmen- tation anchoring.arXiv preprint arXiv:1911.09785, 2019

    David Berthelot, Nicholas Carlini, Ekin D Cubuk, Alex Ku- rakin, Kihyuk Sohn, Han Zhang, and C Raffel. Semi- supervised learning with distribution alignment and augmen- tation anchoring.arXiv preprint arXiv:1911.09785, 2019. 3

    arXiv 1911

  5. [5]

    Mixmatch: A holistic approach to semi-supervised learning.Advances in neural information processing systems, 32, 2019

    David Berthelot, Nicholas Carlini, Ian Goodfellow, Nicolas Papernot, Avital Oliver, and Colin A Raffel. Mixmatch: A holistic approach to semi-supervised learning.Advances in neural information processing systems, 32, 2019. 3

    2019

  6. [6]

    Yolact: Real-time instance segmentation

    Daniel Bolya, Chong Zhou, Fanyi Xiao, and Yong Jae Lee. Yolact: Real-time instance segmentation. InProceedings of the IEEE/CVF international conference on computer vision, pages 9157–9166, 2019. 3

    2019

  7. [7]

    Interactive graph cuts for optimal boundary & region segmentation of objects in nd images

    Yuri Y Boykov and M-P Jolly. Interactive graph cuts for optimal boundary & region segmentation of objects in nd images. InProceedings eighth IEEE international confer- ence on computer vision. ICCV 2001, pages 105–112. IEEE,

    work page 2001

  8. [8]

    Cascade r-cnn: High quality object detection and instance segmentation.IEEE transactions on pattern analysis and machine intelligence, 43(5):1483–1498, 2019

    Zhaowei Cai and Nuno Vasconcelos. Cascade r-cnn: High quality object detection and instance segmentation.IEEE transactions on pattern analysis and machine intelligence, 43(5):1483–1498, 2019. 3

    2019

Show all 46 references
  1. [9]

    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. InEuropean confer- ence on computer vision, pages 213–229. Springer, 2020. 3

    2020

  2. [10]

    Per- pixel classification is not all you need for semantic segmen- tation.Advances in neural information processing systems, 34:17864–17875, 2021

    Bowen Cheng, Alex Schwing, and Alexander Kirillov. Per- pixel classification is not all you need for semantic segmen- tation.Advances in neural information processing systems, 34:17864–17875, 2021. 3

    2021

  3. [11]

    Schwing, Alexan- der Kirillov, and Rohit Girdhar

    Bowen Cheng, Ishan Misra, Alexander G. Schwing, Alexan- der Kirillov, and Rohit Girdhar. Masked-attention mask transformer for universal image segmentation. 2022. 3, 4, 6, 1

    2022

  4. [12]

    Mean shift: A robust ap- proach toward feature space analysis.IEEE Transactions on pattern analysis and machine intelligence, 24(5):603–619,

    Dorin Comaniciu and Peter Meer. Mean shift: A robust ap- proach toward feature space analysis.IEEE Transactions on pattern analysis and machine intelligence, 24(5):603–619,

  5. [13]

    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 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR),...

    2016

  6. [14]

    Semantic instance segmentation with a discriminative loss function.arXiv preprint arXiv:1708.02551, 2017

    Bert De Brabandere, Davy Neven, and Luc Van Gool. Semantic instance segmentation with a discriminative loss function.arXiv preprint arXiv:1708.02551, 2017. 3

    2017 arXiv

  7. [15]

    Improved regular- ization of convolutional neural networks with cutout.arXiv preprint arXiv:1708.04552, 2017

    Terrance DeVries and Graham W Taylor. Improved regular- ization of convolutional neural networks with cutout.arXiv preprint arXiv:1708.04552, 2017. 6, 1

    2017 arXiv

  8. [16]

    Polite teacher: Semi-supervised instance segmentation with mutual learning and pseudo-label thresholding.IEEE Access, 2024

    Dominik Filipiak, Andrzej Zapała, Piotr Tempczyk, Anna Fensel, and Marek Cygan. Polite teacher: Semi-supervised instance segmentation with mutual learning and pseudo-label thresholding.IEEE Access, 2024. 3, 7

    2024

  9. [17]

    Consistency-based semi- supervised active learning: Towards minimizing labeling cost

    Mingfei Gao, Zizhao Zhang, Guo Yu, Sercan ¨O Arık, Larry S Davis, and Tomas Pfister. Consistency-based semi- supervised active learning: Towards minimizing labeling cost. InComputer vision–ECCV 2020: 16th European con- ference, glasgow, UK, August 23–28, 2020, proceedings, par...

    2020

  10. [18]

    Ssap: Single-shot in- stance segmentation with affinity pyramid

    Naiyu Gao, Yanhu Shan, Yupei Wang, Xin Zhao, Yinan Yu, Ming Yang, and Kaiqi Huang. Ssap: Single-shot in- stance segmentation with affinity pyramid. InProceedings of the IEEE/CVF international conference on computer vi- sion, pages 642–651, 2019. 3

    2019

  11. [19]

    Fast r-cnn

    Ross Girshick. Fast r-cnn. InProceedings of the IEEE inter- national conference on computer vision, pages 1440–1448,

  12. [20]

    LVIS: A Dataset for Large V ocabulary Instance Segmentation

    Agrim Gupta, Piotr Doll ´ar, and Ross Girshick. LVIS: A Dataset for Large V ocabulary Instance Segmentation. In IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2019. 1

    2019

  13. [21]

    Pseudo-labeling enhanced by privileged information and its application to in situ sequencing images

    Marzieh Haghighi, Mario C Cruz, Erin Weisbart, Beth A Ci- mini, Avtar Singh, Julia Bauman, Maria E Lozada, Sanam L Kavari, James T Neal, Paul C Blainey, et al. Pseudo-labeling enhanced by privileged information and its application to in situ sequencing images. InProceedings of...

    2023

  14. [22]

    Mask r-cnn

    Kaiming He, Georgia Gkioxari, Piotr Doll ´ar, and Ross Gir- shick. Mask r-cnn. InIEEE International Conference on Computer Vision (ICCV), pages 2961–2969, 2017. 3

    2017

  15. [23]

    Pseudo-label alignment for semi-supervised instance segmentation

    Jie Hu, Chen Chen, Liujuan Cao, Shengchuan Zhang, An- nan Shu, Guannan Jiang, and Rongrong Ji. Pseudo-label alignment for semi-supervised instance segmentation. In 2023 IEEE/CVF International Conference on Computer Vi- sion (ICCV), pages 16291–16301, 2023. 3, 5, 6, 7, 8, 1

    2023

  16. [24]

    Mask scoring r-cnn

    Zhaojin Huang, Lichao Huang, Yongchao Gong, Chang Huang, and Xinggang Wang. Mask scoring r-cnn. InPro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 6409–6418, 2019. 5

    2019

  17. [25]

    Consistency-based semi-supervised learning for object de- tection.Advances in Neural Information Processing Systems,

    Jisoo Jeong, Seungeui Lee, Jeesoo Kim, and Nojun Kwak. Consistency-based semi-supervised learning for object de- tection.Advances in Neural Information Processing Systems,

  18. [26]

    Consistency-based semi-supervised learning for object de- tection.Advances in neural information processing systems, 32, 2019

    Jisoo Jeong, Seungeui Lee, Jeesoo Kim, and Nojun Kwak. Consistency-based semi-supervised learning for object de- tection.Advances in neural information processing systems, 32, 2019. 3

    2019

  19. [27]

    Centermask: Real- time anchor-free instance segmentation

    Youngwan Lee and Jongyoul Park. Centermask: Real- time anchor-free instance segmentation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 13906–13915, 2020. 3

    2020

  20. [28]

    Exploring plain vision transformer backbones for object de- tection

    Yanghao Li, Hanzi Mao, Ross Girshick, and Kaiming He. Exploring plain vision transformer backbones for object de- tection. InEuropean conference on computer vision, pages 280–296. Springer, 2022. 3

    2022

  21. [29]

    Weakly supervised open- vocabulary object detection

    Jianghang Lin, Yunhang Shen, Bingquan Wang, Shaohui Lin, Ke Li, and Liujuan Cao. Weakly supervised open- vocabulary object detection. InProceedings of the AAAI Conference on Artificial Intelligence, pages 3404–3412,

  22. [30]

    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 European Conference on Computer Vision (ECCV), 2014. 1, 6

    2014

  23. [31]

    Sgn: Sequential grouping networks for instance segmentation

    Shu Liu, Jiaya Jia, Sanja Fidler, and Raquel Urtasun. Sgn: Sequential grouping networks for instance segmentation. In Proceedings of the IEEE international conference on com- puter vision, pages 3496–3504, 2017. 3

    2017

  24. [32]

    Unbiased teacher for semi-supervised object detec- tion.International Conference on Learning Representations,

    Yen-Cheng Liu, Chih-Yao Ma, Zijian He, Chia-Wen Kuo, Kan Chen, Peizhao Zhang, Bichen Wu, Zsolt Kira, and Peter Vajda. Unbiased teacher for semi-supervised object detec- tion.International Conference on Learning Representations,

  25. [33]

    Decoupled weight decay regularization.arXiv preprint arXiv:1711.05101, 2017

    Ilya Loshchilov and Frank Hutter. Decoupled weight decay regularization.arXiv preprint arXiv:1711.05101, 2017. 6, 1

    2017 arXiv

  26. [34]

    Semi-supervised se- mantic segmentation via strong-weak dual-branch network

    Wenfeng Luo and Meng Yang. Semi-supervised se- mantic segmentation via strong-weak dual-branch network. Springer International Publishing eBooks, 2020. 7

    2020

  27. [35]

    Active teacher for semi-supervised ob- ject detection

    Peng Mi, Jianghang Lin, Yiyi Zhou, Yunhang Shen, Gen Luo, Xiaoshuai Sun, Liujuan Cao, Rongrong Fu, Qiang Xu, and Rongrong Ji. Active teacher for semi-supervised ob- ject detection. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 14482...

    2022

  28. [36]

    Semi- supervised semantic segmentation with cross-consistency training.arXiv: Computer Vision and Pattern Recognition,

    Yassine Ouali, C ´eline Hudelot, and Myriam Tami. Semi- supervised semantic segmentation with cross-consistency training.arXiv: Computer Vision and Pattern Recognition,

  29. [37]

    Learning transferable visual models from natural language supervi- sion

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learning transferable visual models from natural language supervi- sion. InInternational conference on machine learning, p...

    2021

  30. [38]

    Girshick, Georgia Gkioxari, and Kaiming He

    Ilija Radosavovic, Piotr Doll ´ar, Ross B. Girshick, Georgia Gkioxari, and Kaiming He. Data distillation: Towards omni- supervised learning. Incvpr, 2018. 7

    2018

  31. [39]

    A simple semi-supervised learning framework for object detection.arXiv preprint arXiv:2005.04757, 2020

    Kihyuk Sohn, Zizhao Zhang, Chun-Liang Li, Han Zhang, Chen-Yu Lee, and Tomas Pfister. A simple semi-supervised learning framework for object detection.arXiv preprint arXiv:2005.04757, 2020. 3, 7

    2005 arXiv

  32. [40]

    Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results.Advances in neural information processing systems, 30, 2017

    Antti Tarvainen and Harri Valpola. Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results.Advances in neural information processing systems, 30, 2017. 4

    2017

  33. [41]

    Noisy bound- aries: Lemon or lemonade for semi-supervised instance seg- mentation?2022 IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition (CVPR), pages 16805–16814,

    Zhenyu Wang, Yali Li, and Shengjin Wang. Noisy bound- aries: Lemon or lemonade for semi-supervised instance seg- mentation?2022 IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition (CVPR), pages 16805–16814,

    2022

  34. [42]

    Detectron2.https://github

    Yuxin Wu, Alexander Kirillov, Francisco Massa, Wan-Yen Lo, and Ross Girshick. Detectron2.https://github. com/facebookresearch/detectron2, 2019. 6, 1

    2019

  35. [43]

    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 Proceedings of the IEEE/CVF international conference on computer vision, pages 3060–3069, 2021. 3

    2021

  36. [44]

    Detclipv2: Scal- able open-vocabulary object detection pre-training via word- region alignment

    Lewei Yao, Jianhua Han, Xiaodan Liang, Dan Xu, Wei Zhang, Zhenguo Li, and Hang Xu. Detclipv2: Scal- able open-vocabulary object detection pre-training via word- region alignment. InProceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pages 23497...

    2023

  37. [45]

    Bdd100k: A diverse driving dataset for heterogeneous multitask learning

    Fisher Yu, Haofeng Chen, Xin Wang, Wenqi Xian, Yingying Chen, Fangchen Liu, Vashisht Madhavan, and Trevor Dar- rell. Bdd100k: A diverse driving dataset for heterogeneous multitask learning. InIEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 2636–2645,

  38. [46]

    burn-in” stage to train our models on only labeled data. After that, run a teacher-student “mutual learning

    Xingyi Zhou, Rohit Girdhar, Armand Joulin, Philipp Kr¨ahenb¨uhl, and Ishan Misra. Detecting twenty-thousand classes using image-level supervision. InEuropean Confer- ence on Computer Vision, pages 350–368. Springer, 2022. 5 Pseudo-Label Quality Decoupling and Correction for Se...

    2022

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.