Pith. sign in

REVIEW 4 major objections 7 minor 66 references

Stable Mean Teacher for Semi-supervised Video Action Detection

T0 review · 4 major / 7 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Stable Mean Teacher claims that semi-supervised video action detection reaches near fully-supervised quality with only 10–20% of per-frame labels.

desk verdict Solid empirical extension of Mean Teacher to spatio-temporal action detection with two useful modules, but the headline '10%/20% vs 100%' comparison is not anchored by a same-base fully supervised run. read the letter →

arxiv 2412.07072 v2 pith:3BHAFNQA submitted 2024-12-10 cs.CV

Akash Kumar , Sirshapan Mitra , Yogesh Singh Rawat This is my paper · ORCID
classification cs.CV
keywords semi-supervisedlearningvideoactiondetectionmeanteacherErrorRecoverymoduleDifferenceofPixelstemporalconsistencypseudo-labelslabelefficiency
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

Stable Mean Teacher claims that semi-supervised video action detection can be made much more label-efficient with a student–teacher scheme that upgrades pseudo-labels in two concrete ways: a class-agnostic Error Recovery module that refines the teacher's spatial localization maps, and a Difference of Pixels constraint that keeps those maps temporally coherent. On UCF101-24, training with 10% of the frame-level annotations reaches 73.9 f-mAP@0.5 and 76.3 v-mAP@0.5, which the paper reports as competitive with fully supervised training at 100% labels and about 20 and 27 points above a supervised baseline trained on the same 10% subset. On JHMDB21, 20% labels give 69.8 f-mAP@0.5 and 70.7 v-mAP@0.5, also reported as competitive with full supervision. Reported margins over the same-percentage supervised baseline average 23.5% on UCF101-24 and 16% on JHMDB21. If these numbers hold, per-frame annotation cost for video action detection could drop by an order of magnitude while keeping most of the quality.

What carries the argument

Two pseudo-label refinements carry the argument. Error Recovery (EoR) is a class-agnostic 3D U-Net that sees only the teacher's localization map, not the input video or class labels, so it learns a boundary-correcting transformation that can generalize to unlabeled clips; the teacher-side EoR weights follow an EMA of the student-side EoR, and gradient stopping keeps the base model unchanged. Difference of Pixels (DoP) computes $\phi(x^f_{loc}) = x^{f+1}_{loc} - x^f_{loc}$ for each pair of consecutive frames and enforces MSE consistency between teacher and student on these temporal differences, pushing predictions to move coherently across time. The full objective (Eq. 7) is $L = L_s + \lambda (L_{base} + L_{EoR} + L_{DoP})$, with $\lambda = 0.1$ and EMA decay $\beta = 0.99$.

What would settle it

On a held-out subset of UCF101-24 at the 10% labeled setting, measure the IoU between teacher pseudo-labels and ground-truth boxes on unlabeled videos before and after EoR refinement; if the refined pseudo-labels are not systematically closer to ground truth, the claimed error-recovery mechanism is not doing the work attributed to it.

Watch

Extended reading notes

Core claim

The paper's central claim is that a Mean Teacher framework adapted to spatiotemporal detection, called Stable Mean Teacher, can generate pseudo-labels good enough that a small labeled subset yields near-fully-supervised action detection quality. The teacher is an exponential-moving-average of the student and sees weakly augmented clips, while the student sees strongly augmented clips; consistency between them is the unsupervised training signal. The Error Recovery (EoR) module, a class-agnostic 3D U-Net with roughly 1.1M parameters, takes the teacher's raw localization map and produces a refined map; its loss (Eq. 4) is the MSE between the refined teacher map and the student's localization output, with gradients intentionally stopped from updating the base model. The Difference of Pixels (DoP) constraint (Eqs. 5–6) enforces agreement on the temporal difference of consecutive localization maps, making pseudo-labels temporally coherent. The paper reports that on UCF101-24, JHMDB21, and AVA this framework outperforms the same-scale supervised baseline by average margins of 23.5%, 16%, and 3.3%, respectively, and that at 10–20% annotations it approaches fully supervised performance.

Load-bearing premise

The load-bearing assumption is that the teacher's pseudo-labels, and the EoR module's refinements of them, are accurate enough to serve as training targets: the method applies no confidence threshold or noise filtering, and the EoR objective in Eq. 4 is a self-consistency loss with no ground-truth error signal, so confident systematic teacher mistakes could be reinforced instead of corrected.

Editorial extensions

If this is right

  • With 10% of UCF101-24 labels, Stable Mean Teacher reaches f-mAP@0.5 of 73.9 and v-mAP@0.5 of 76.3, outperforming the 10%-supervised baseline by about 20.4 and 26.6 points.
  • With 20% of JHMDB21 labels, it reaches f-mAP@0.5 of 69.8 and v-mAP@0.5 of 70.7, outperforming the 20%-supervised baseline by about 14.1 and 18.3 points.
  • It beats the prior semi-supervised video action detection baseline by 8.7 f-mAP@0.5 and 9.6 v-mAP@0.5 on UCF101-24, and by 5.4 and 7.2 on JHMDB21, the latter with 10% less labeled data.
  • On AVA, the method improves a real-time detector's mAP by 3.3 points using only 10% labels, indicating it scales to a large dataset.
  • On YouTube-VOS, 10% labels give an average Jaccard of 41.3, above the tested semi-supervised baseline and close to the fully supervised 47.9, evidence that the approach transfers to video object segmentation.

Reading between the lines

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

  • A natural next experiment the paper does not run is to add confidence-based pseudo-label selection on top of EoR/DoP; if teacher confidence correlates with correctness, filtering could push the low-label regime even higher.
  • The large gains on dynamic scenes (39.4% at v-mAP@0.5 on JHMDB21 dynamic classes) suggest the DoP constraint is the main driver where background and actor motion are strong; this could be tested by ablating DoP separately on static and dynamic class splits.
  • Because the EoR module is class-agnostic and the teacher-student recipe is detector-agnostic, the same pseudo-label refinement may transfer to other per-frame annotation-heavy video tasks, such as video instance segmentation or multi-person pose tracking; the YouTube-VOS result is preliminary but consistent with that.
  • The method's gains grow as labels get sparser (on UCF101-24 the relative gain at 5% labels is roughly double the gain at 10%), which suggests the approach is most valuable precisely when labels are hardest to obtain.
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

4 major / 7 minor

Summary. The paper proposes Stable Mean Teacher, an EMA student-teacher framework for semi-supervised video action detection. It introduces two components: an Error Recovery (EoR) module, a 3D U-Net that refines the teacher's localization outputs, and a Difference of Pixels (DoP) constraint that enforces temporal consistency on localization maps. The method is evaluated on UCF101-24, JHMDB21, AVA, and YouTube-VOS, with claims that 10% labels on UCF101-24 and 20% labels on JHMDB21 are competitive with fully supervised training, and that large gains are obtained over same-label supervised baselines.

Significance. If the claims hold, the paper would be a useful step toward label-efficient video action detection. The work has clear strengths: code and models are publicly available; the ablations in Table 3 and supplementary Table 9 isolate the contributions of EoR and DoP; and the evaluation spans four benchmarks with multiple thresholds and a generalization experiment to video object segmentation. The EoR idea of a class-agnostic correction module trained with a stopped gradient is interesting and the overall results are plausible. However, the headline comparison to full supervision is not anchored in the main tables, and the EoR training objective as written does not contain the ground-truth error signal that the text claims. These points must be resolved before the central claims are accepted.

major comments (4)
  1. [Table 1; Abstract; Figure 1] The abstract's claim that 10% and 20% labels give performance competitive with a supervised baseline trained on 100% annotations is not anchored by a same-pipeline fully supervised run in the main text. The rows labeled 'Supervised baseline' in Tables 1, 5, 6, and 7 are trained on the labeled subset (10% or 20%), not on 100%. The supplementary Table 8a does report '100% sup.' numbers for a static/dynamic split, but no aggregate full-supervision number for the exact VideoCapsuleNet/STMT pipeline appears in the main comparison. Please add a same-model, same-recipe 100%-supervised row to the comparison tables, report the precise gaps that support the word 'competitive,' and adjust the abstract or figure accordingly if those gaps are larger than implied.
  2. [Methodology, Error Recovery (EoR), Eq. (4)] The EoR loss in Eq. (4) is written as LEoR = MSE(MEoR_t(tloc), sloc), which is a teacher-student consistency loss with no ground-truth term. The text states that EoR 'learns from students' mistakes on labeled samples,' but Eq. (4) contains no supervision from labeled annotations. If the supervised loss Ls is also applied to the EoR module's output on labeled samples, as suggested by the sentence 'and student's Error Recovery module predictions,' that term is not written in Eq. (7) or the surrounding text. Please specify the complete EoR objective on labeled samples, or explain how Eq. (4) alone can teach the module to recover true errors rather than simply imitate the teacher. Without this clarification, the 'error recovery' interpretation is not supported by the equations.
  3. [Abstract; Comparison with supervised and weakly-supervised] The statements 'outperforms the supervised baselines by 23.5%' and 'competitive performance compared to the supervised baseline trained on 100% annotations' conflate two different baselines. The 23.5% margin is computed against the same-label (10% or 20%) supervised baseline in Table 1, not against a fully supervised model. Please rename the rows (e.g., 'Supervised (same label fraction)') and state explicitly which baseline is used for each quantitative claim, so that the abstract's wording does not mislead readers into thinking the 23.5% gain is measured against full supervision.
  4. [Tables 1-9] No error bars or multiple-seed results are reported. Several of the claimed improvements are small, notably the DoP gains of 0.5-1% in Table 3 and the EoR gain of about 1% on UCF101-24 f-mAP@0.5 in the same table. Without variance estimates or a statement of the number of seeds, these differences could be run-to-run noise. Please report mean and standard deviation over at least three seeds for the key comparisons in Tables 1, 3, and 5, or explicitly state that the reported numbers are already averages over multiple runs.
minor comments (7)
  1. [Eq. (4)] The notation 'M SE' in Eq. (4) should be typeset as a single operator, e.g., \operatorname{MSE}.
  2. [Methodology, Learning objectives] 'Jenson-Shennon Divergence' should be 'Jensen-Shannon Divergence.'
  3. [Abstract] The word 'psuedo-labels' in the abstract should be 'pseudo-labels.'
  4. [Table 2] The column headers K, FPS, A, mAP, and GFLOPs are not all defined in the caption; please define K and A explicitly.
  5. [Figure 1] The figure legend appears to include a 'Ours 100%' entry, but Table 1 contains no 100%-supervised row for the authors' pipeline; please align the figure with the new baseline row requested in the major comments.
  6. [Comparison with supervised and weakly-supervised] The comparison with fully supervised methods that use different backbones and modalities (e.g., YOWO with ResNext-101, TACNet with optical flow) should be caveated more explicitly; the wording 'surpasses all the 2D-based approaches' could be misread as a same-architecture comparison.
  7. [Table 4] In the VOS results, the sentence 'outperforms the supervised baseline by an absolute margin of 31%' should identify which row is the 'supervised baseline' (it appears to be Xu 10%), and clarify that the comparison is not against the Xu 100% row.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: the EMA/EoR self-consistency loop is the proposed method, and the reported gains are measured on fixed external benchmarks rather than being forced by construction.

full rationale

The paper's central empirical claim is evaluated on established test splits (UCF101-24, JHMDB21, AVA, YouTube-VOS) against published numbers, so the headline improvement is not a fitted artifact of the derivation. The only self-referential element is the standard Mean Teacher EMA loop (Eqs. 1 and 3) and the consistency losses in Eqs. 4 and 5, where the teacher target is an EMA of the student and the student is trained to match it. This is the method itself, not a hidden reuse of the evaluation target; such bootstrap-style training can genuinely improve performance and is not equivalent to the input by construction. The EoR module also receives supervised signal on labeled samples according to the text ('We calculate the supervised loss on the labeled subset of student's predictions ... and student's Error Recovery module predictions'), which provides an external error signal for the 'error recovery' interpretation. Ablations (Table 3 and Table 9) isolate the contributions of EoR and DoP, and the unsupervised weight lambda is set empirically rather than fitted to the reported test metrics. Self-citations such as Kumar and Rawat (2022) supply the base architecture and a comparison point, but they do not carry the argument: the method is tested against independent published methods and fixed benchmarks. A separate weakness is that the abstract's 'trained on 100% annotations' comparison is not backed by a same-pipeline fully supervised run, but a missing baseline is an evidence/comparison concern, not circularity.

Assumptions & free parameters 4 free parameters · 4 assumptions · 2 invented entities

The central empirical claims rest on standard SSL assumptions plus design choices: the paper assumes unlabeled data helps, assumes teacher pseudo-labels can be trusted without filtering, assumes the EoR transformation transfers across samples, and assumes temporal smoothness is a useful prior. Hyperparameters lambda and beta are hand-set, and the EoR architecture and augmentation schedule are chosen by the authors. No new physical entities are postulated; EoR and DoP are pipeline components whose only evidence is internal ablations.

free parameters (4)
  • Unsupervised loss weight lambda = 0.1
    Set empirically in the implementation details. It weights all unsupervised losses in Eq. 7 and directly controls the contribution of pseudo-label consistency.
  • EMA decay beta = 0.99
    Set by hand following prior works. Controls how slowly the teacher tracks the student and is central to pseudo-label generation.
  • EoR 3D UNet capacity = 16, 32, 64, 128 channels, about 1.1M parameters
    Architecture capacity is chosen by the authors. It is a design choice that affects the claimed error-recovery behavior.
  • Strong augmentation probabilities and magnitudes = Contrast/Hue/Brightness/Saturation 0.7, Grayscale 0.6, Gaussian Blur 0.5
    Hand-picked augmentation schedule in the supplementary defines the student-teacher view pair. These choices affect the pseudo-label quality.
assumptions (4)
  • domain assumption Unlabeled and labeled videos are drawn from the same distribution, so teacher pseudo-labels on unlabeled data provide a valid training signal.
    Standard SSL premise used in Section 'Problem formulation'; no mechanism handles distribution shift or noisy labels.
  • domain assumption EMA teacher pseudo-labels are reliable enough to be used as targets without confidence filtering.
    All unsupervised losses in Eq. 7 are applied to every unlabeled sample; confidently wrong teacher predictions can reinforce their own errors.
  • domain assumption The error transformation learned by EoR on labeled samples generalizes to unlabeled samples.
    Stated in the 'Gradient flow' paragraph; no analysis of error distribution or out-of-distribution generalization is provided.
  • domain assumption Minimizing temporal differences of localization maps improves action detection quality.
    DoP loss is justified by intuition about temporal coherency; the paper tests it only indirectly through final mAP benchmarks.
invented entities (2)
  • Error Recovery (EoR) module
    purpose: A 3D UNet that takes the base model localization map and outputs a refined localization map. Claimed to learn and correct student localization errors and improve teacher pseudo-labels.
    The only evidence for its claimed error-recovery function is the paper's own ablations and qualitative figures. There is no external falsifiable prediction, and Eq. 4 is a consistency loss rather than a labeled-error recovery loss.
  • Difference of Pixels (DoP) constraint
    purpose: A temporal-consistency loss that minimizes the MSE between consecutive-frame differences of student and teacher localization maps.
    The constraint exists only inside this pipeline. Its benefit is shown through internal ablations, not through an independent measurement or benchmark specifically designed to test temporal coherence.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Stable Mean Teacher for Semi-supervised Video Action Detection." pith.science (2026). https://pith.science/paper/3BHAFNQA

@misc{pith2026241207072,
  author       = {Pith},
  title        = {Pith review of: Stable Mean Teacher for Semi-supervised Video Action Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3BHAFNQA}},
  note         = {Machine review of arXiv:2412.07072}
}
read the original abstract

In this work, we focus on semi-supervised learning for video action detection. Video action detection requires spatiotemporal localization in addition to classification, and a limited amount of labels makes the model prone to unreliable predictions. We present Stable Mean Teacher, a simple end-to-end teacher-based framework that benefits from improved and temporally consistent pseudo labels. It relies on a novel Error Recovery (EoR) module, which learns from students' mistakes on labeled samples and transfers this knowledge to the teacher to improve pseudo labels for unlabeled samples. Moreover, existing spatiotemporal losses do not take temporal coherency into account and are prone to temporal inconsistencies. To address this, we present Difference of Pixels (DoP), a simple and novel constraint focused on temporal consistency, leading to coherent temporal detections. We evaluate our approach on four different spatiotemporal detection benchmarks: UCF101-24, JHMDB21, AVA, and YouTube-VOS. Our approach outperforms the supervised baselines for action detection by an average margin of 23.5% on UCF101-24, 16% on JHMDB21, and 3.3% on AVA. Using merely 10% and 20% of data, it provides competitive performance compared to the supervised baseline trained on 100% annotations on UCF101-24 and JHMDB21, respectively. We further evaluate its effectiveness on AVA for scaling to large-scale datasets and YouTube-VOS for video object segmentation, demonstrating its generalization capability to other tasks in the video domain. Code and models are publicly available.

Figures

Figures reproduced from arXiv: 2412.07072 by the authors.

Figure 1
Figure 1. Performance overview: Stable Mean Teacher provides comparable performance with 10% (UCF101-24; left two plots) and 20% (JHMDB-21; right two plots) labels when compared with fully supervised approach which is trained on 100% annotations. It consistently outperforms existing state-of-the-art (2022) and supervised baseline on both f-mAP and v-mAP with good margin on both UCF101-24 and JHMDB-21 at all different percenta… view at source ↗
Figure 2
Figure 2. Overview of Stable Mean Teacher. The two key components to improve the quality of spatio-temporal pseudo label: 1) Error Recovery: refines the spatial action boundary, 2) DoP constraint: induces temporal coherency on predicted spatio-temporal pseudo labels. Different from all these, we focus on videos where the temporal dimension adds more complexity to the problem. There are some recent works focusing on videos, bu… view at source ↗
Figure 3
Figure 3. Visualization of Difference of Pixels (DoP). First row shows the RGB frames, second row shows the pixel difference map of ground truth along temporal dimension. We show two scenarios: Left: Static: constant background; actor in motion, and Right: Dynamic: changing background; actor in motion. Temporal difference emphasizes on the variation of boundary pixels between consecutive frames. available per class. This limi… view at source ↗
Figures from the paper (10 more)
Figure 4
Figure 4. Figure 4: Qualitative analysis for EoR and DoP: Left side illustrates the effectiveness of Error Recovery module on multiple samples, with improvement in action boundary pre￾cision and it also helps in suppressing background noise. On the right hand, we demonstrate how DoP const…
Figure 5
Figure 5. Figure 5: Analyzing Stable Mean Teacher: (Left) Static vs dynamic scenes: Dynamic scenes are challenging than static scenes, however, the relative boost in performance for dynamic is 27.7% more than in case of static scene scenario. ∆ denotes relative change at v-mAP@0.5. (Middl…
Figure 6
Figure 6. Figure 6: Classwise analysis: Improvement in v-mAP@0.5 for top 3 action classes with maximum performance gain over supervised baseline on static: {throw, sit, brushhair} and dynamic {diving, skating, surfing} showing effectiveness of the proposed ap￾proach. Method Annot. Avg JS …
Figure 7
Figure 7. Figure 7: This figure shows the top 5 classes which has the most improvement on v-mAP@0.5 on our proposed semi-supervised [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]
Figure 8
Figure 8. Figure 8: This figure shows the top 5 classes which has the most improvement on v-mAP@0.5 on our proposed semi-supervised [PITH_FULL_IMAGE:figures/full_fig_p011_8.png]
Figure 9
Figure 9. Figure 9: Visualization of augmentations: This figure shows the original clip and augmented clip from UCF101 and JH￾MDB21 dataset respectively. tained. 2) Temporal: This augmentation is also similar for both teacher and student. This is because we need to cal￾culate localization…
Figure 10
Figure 10. Figure 10: Qualitative results - Case - I - Boundary Refinement In this scenario, we can see the Ours could even separate out the instance of two legs separately which shows that the precise error from EoR model helps in refinement for fine-grained details. The predictions are e…
Figure 11
Figure 11. Figure 11: Qualitative results - Case - II - Noise Suppression In this scenario, we can see the Ours is able to suppress the background noise more better. The detaching of EoR module from main model helps this procedure. Otherwise, the mispredic￾tions gets enhanced. Jing, L.; Pa…
Figure 12
Figure 12. Figure 12: Qualitative results - Case - III - Noise Suppression + Boundary Refinement In this scenario, model is able to do both getting rid of noise and refining the boundary at the same time. Even 100% supervised model fails at it. Liu, Y.-C.; Ma, C.-Y.; He, Z.; Kuo, C.-W.; Ch…
Figure 13
Figure 13. Figure 13: Qualitative results - Case - IV - Temporal Mask Coherency In this scenario, we show that Ours not only helps to localize the actor spatially but the temporal coherency of mask is also maintained in case of large displacement/motion. Ren, S.; He, K.; Girshick, R.; and …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

66 extracted references · 50 canonical work pages

  1. [1]

    Arnab, A.; Sun, C.; Nagrani, A.; and Schmid, C. 2020. Uncertainty-Aware Weakly Supervised Action Detection from Untrimmed Videos. ArXiv, abs/2007.10703

    work page Pith review arXiv 2020

  2. [2]

    Berthelot, D.; Carlini, N.; Goodfellow, I.; Papernot, N.; Oliver, A.; and Raffel, C. A. 2019. MixMatch: A Holistic Approach to Semi-Supervised Learning. In Wallach, H.; Larochelle, H.; Beygelzimer, A.; d Alch\' e -Buc, F.; Fox, E.; and Garnett, R., eds., Advances in Neural Information Processing Systems, volume 32. Curran Associates, Inc

    work page 2019

  3. [3]

    Carion, N.; Massa, F.; Synnaeve, G.; Usunier, N.; Kirillov, A.; and Zagoruyko, S. 2020. End-to-End Object Detection with Transformers. ArXiv, abs/2005.12872

    arXiv 2020

  4. [4]

    Chen, B.; Li, P.; Chen, X.; Wang, B.; Zhang, L.; and Hua, X.-S. 2022. Dense Learning Based Semi-Supervised Object Detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 4815--4824

    work page 2022

  5. [5]

    Chen, L.; Tong, Z.; Song, Y.; Wu, G.; and Wang, L. 2023. Efficient video action detection with token dropout and context refinement. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 10388--10399

    2023

  6. [6]

    Chen, S.; Sun, P.; Xie, E.; Ge, C.; Wu, J.; Ma, L.; Shen, J.; and Luo, P. 2021. Watch only once: An end-to-end video action detection framework. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 8178--8187

    2021

  7. [7]

    Ch\' e ron, G.; Alayrac, J.-B.; Laptev, I.; and Schmid, C. 2018. A flexible model for training action localization with varying levels of supervision. In Bengio, S.; Wallach, H.; Larochelle, H.; Grauman, K.; Cesa-Bianchi, N.; and Garnett, R., eds., Advances in Neural Information Processing Systems, volume 31. Curran Associates, Inc

    work page 2018

  8. [8]

    S.; and Shah, M

    Dave, I.; Scheffer, Z.; Kumar, A.; Shiraz, S.; Rawat, Y. S.; and Shah, M. 2022. GabriellaV2: Towards Better Generalization in Surveillance Videos for Action Detection. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) Workshops, 122--132

    work page 2022

Show all 66 references
  1. [9]

    S.; and Shah, M

    Duarte, K.; Rawat, Y. S.; and Shah, M. 2018. Videocapsulenet: A simplified network for action detection. Advances in Neural Information Processing Systems

    2018

  2. [10]

    D.; Jain, M.; Ghanem, B.; and Snoek, C

    Escorcia, V.; Dao, C. D.; Jain, M.; Ghanem, B.; and Snoek, C. G. M. 2020. Guess Where? Actor-Supervision for Spatiotemporal Action Localization. Comput. Vis. Image Underst., 192: 102886

    2020

  3. [11]

    feng Zhou, Q.; Yu, C.; Wang, Z.; Qian, Q.; and Li, H. 2021. Instant-Teaching: An End-to-End Semi-Supervised Object Detection Framework. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 4079--4088

    2021

  4. [12]

    Finn, C.; Goodfellow, I.; and Levine, S. 2016. Unsupervised learning for physical interaction through video prediction. arXiv preprint arXiv:1605.07157

    2016 arXiv

  5. [13]

    Gkioxari, G.; Girshick, R.; Doll \'a r, P.; and He, K. 2018. Detecting and recognizing human-object interactions. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 8359--8367

    2018

  6. [14]

    A.; Vondrick, C.; Pantofaru, C.; Li, Y.; Vijayanarasimhan, S.; Toderici, G.; Ricco, S.; Sukthankar, R.; Schmid, C.; and Malik, J

    Gu, C.; Sun, C.; Ross, D. A.; Vondrick, C.; Pantofaru, C.; Li, Y.; Vijayanarasimhan, S.; Toderici, G.; Ricco, S.; Sukthankar, R.; Schmid, C.; and Malik, J. 2018. AVA: A Video Dataset of Spatio-Temporally Localized Atomic Visual Actions. In 2018 IEEE/CVF Conference on Computer ...

    2018

  7. [15]

    Hou, R.; Chen, C.; and Shah, M. 2017. Tube convolutional neural network (T-CNN) for action detection in videos. In IEEE International Conference on Computer Vision

    2017

  8. [16]

    Jeong, J.; Verma, V.; Hyun, M.; Kannala, J.; and Kwak, N. 2021. Interpolation-based Semi-supervised Learning for Object Detection. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 11597--11606

    2021

  9. [17]

    Jhuang, H.; Gall, J.; Zuffi, S.; Schmid, C.; and Black, M. J. 2013. Towards understanding action recognition. In International Conf. on Computer Vision (ICCV), 3192--3199

    2013

  10. [18]

    Ji, J.; Cao, K.; and Niebles, J. C. 2019. Learning Temporal Action Proposals With Fewer Labels. 2019 IEEE/CVF International Conference on Computer Vision (ICCV), 7072--7081

    2019

  11. [19]

    Jing, L.; Parag, T.; Wu, Z.; Tian, Y.; and Wang, H. 2021. VideoSSL: Semi-Supervised Learning for Video Classification. 2021 IEEE Winter Conference on Applications of Computer Vision (WACV), 1109--1118

    2021

  12. [20]

    Ke, Z.; Wang, D.; Yan, Q.; Ren, J. S. J.; and Lau, R. W. H. 2019. Dual Student: Breaking the Limits of the Teacher in Semi-Supervised Learning. 2019 IEEE/CVF International Conference on Computer Vision (ICCV), 6727--6735

    2019

  13. [21]

    o p \"u kl \

    K \"o p \"u kl \"u , O.; Wei, X.; and Rigoll, G. 2019. You only watch once: A unified cnn architecture for real-time spatiotemporal action localization. arXiv preprint arXiv:1911.06644

    2019 arXiv

  14. [22]

    Kumar, A.; Kumar, A.; Vineet, V.; and Rawat, Y. S. 2023. Benchmarking self-supervised video representation learning. arXiv preprint arXiv:2306.06010

    2023 arXiv

  15. [23]

    Kumar, A.; and Rawat, Y. S. 2022. End-to-End Semi-Supervised Learning for Video Action Detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

    2022

  16. [24]

    Laine, S.; and Aila, T. 2017. Temporal Ensembling for Semi-Supervised Learning. ArXiv, abs/1610.02242

    2017 arXiv

  17. [25]

    Lee, D.-H.; et al. 2013. Pseudo-label: The simple and efficient semi-supervised learning method for deep neural networks. In Workshop on challenges in representation learning, ICML, volume 3, 896

    2013

  18. [26]

    Li, Y.; Wang, Z.; Wang, L.; and Wu, G. 2020. Actions as Moving Points. In arXiv preprint arXiv:2001.04608

    2020 arXiv

  19. [27]

    E.; Fu, C.-Y.; and Berg, A

    Liu, W.; Anguelov, D.; Erhan, D.; Szegedy, C.; Reed, S. E.; Fu, C.-Y.; and Berg, A. 2016. SSD: Single Shot MultiBox Detector. In ECCV

    2016

  20. [28]

    Liu, Y.-C.; Ma, C.-Y.; Dai, X.; Tian, J.; Vajda, P.; He, Z.; and Kira, Z. 2022. Open-Set Semi-Supervised Object Detection. In European Conference on Computer Vision

    2022

  21. [29]

    Liu, Y.-C.; Ma, C.-Y.; He, Z.; Kuo, C.-W.; Chen, K.; Zhang, P.; Wu, B.; Kira, Z.; and Vajda, P. 2021. Unbiased Teacher for Semi-Supervised Object Detection. In Proceedings of the International Conference on Learning Representations (ICLR)

    2021

  22. [30]

    Liu, Y.-C.; Ma, C.-Y.; and Kira, Z. 2022. Unbiased Teacher v2: Semi-Supervised Object Detection for Anchor-Free and Anchor-Based Detectors. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 9819--9828

    2022

  23. [31]

    Mettes, P.; and Snoek, C. G. M. 2018. Pointly-Supervised Action Localization. International Journal of Computer Vision, 127: 263--281

    2018

  24. [32]

    Mettes, P.; Snoek, C. G. M.; and Chang, S.-F. 2017. Localizing Actions from Video Labels and Pseudo-Annotations. ArXiv, abs/1707.09143

    2017 arXiv

  25. [33]

    Nag, S.; Zhu, X.; Song, Y.-Z.; and Xiang, T. 2022. Semi-Supervised Temporal Action Detection with Proposal-Free Masking. In European Conference on Computer Vision

    2022

  26. [34]

    Ntinou, I.; Sanchez, E.; and Tzimiropoulos, G. 2024. Multiscale vision transformers meet bipartite matching for efficient single-stage action localization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 18827--18836

    2024

  27. [35]

    Z.; Liu, Y.; Shao, J.; and Li, H

    Pan, J.; Chen, S.; Shou, M. Z.; Liu, Y.; Shao, J.; and Li, H. 2021. Actor-context-actor relation network for spatio-temporal action localization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 464--474

    2021

  28. [36]

    Pham, H.; Dai, Z.; Xie, Q.; and Le, Q. V. 2021. Meta pseudo labels. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 11557--11568

    2021

  29. [37]

    Rasmus, A.; Valpola, H.; Honkala, M.; Berglund, M.; and Raiko, T. 2015. Semi-Supervised Learning with Ladder Network. ArXiv, abs/1507.02672

    2015 arXiv

  30. [38]

    Ren, S.; He, K.; Girshick, R.; and Sun, J. 2015. Faster r-cnn: Towards real-time object detection with region proposal networks. In Advances in neural information processing systems, 91--99

    2015

  31. [39]

    N.; Duarte, K.; Rawat, Y

    Rizve, M. N.; Duarte, K.; Rawat, Y. S.; and Shah, M. 2020. In Defense of Pseudo-Labeling: An Uncertainty-Aware Pseudo-label Selection Framework for Semi-Supervised Learning. In International Conference on Learning Representations

    2020

  32. [40]

    Ronneberger, O.; Fischer, P.; and Brox, T. 2015. U-Net: Convolutional Networks for Biomedical Image Segmentation. ArXiv, abs/1505.04597

    2015 arXiv

  33. [41]

    Sajjadi, M. S. M.; Javanmardi, M.; and Tasdizen, T. 2016. Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learning. In NIPS

    2016

  34. [42]

    S.; Saenko, K.; and Das, A

    Singh, A.; Chakraborty, O.; Varshney, A.; Panda, R.; Feris, R. S.; Saenko, K.; and Das, A. 2021. Semi-Supervised Action Recognition with Temporal Contrastive Learning. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 10384--10394

    2021

  35. [43]

    J.; Kumar, A.; Vyas, S.; and Rawat, Y

    Singh, A.; Rana, A. J.; Kumar, A.; Vyas, S.; and Rawat, Y. S. 2024. Semi-supervised Active Learning for Video Action Detection. Proceedings of the AAAI Conference on Artificial Intelligence, 38(5): 4891--4899

    2024

  36. [44]

    A.; Cubuk, E

    Sohn, K.; Berthelot, D.; Carlini, N.; Zhang, Z.; Zhang, H.; Raffel, C. A.; Cubuk, E. D.; Kurakin, A.; and Li, C.-L. 2020. FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence. In Larochelle, H.; Ranzato, M.; Hadsell, R.; Balcan, M. F.; and Lin, H., ed...

    2020

  37. [45]

    Song, L.; Zhang, S.; Yu, G.; and Sun, H. 2019. TACNet: Transition-Aware Context Network for Spatio-Temporal Action Detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

    2019

  38. [46]

    Soomro, K.; Zamir, A.; and Shah, M. 2012. UCF101: A Dataset of 101 Human Actions Classes From Videos in The Wild. ArXiv, abs/1212.0402

    2012 arXiv

  39. [47]

    Sui, L.; Zhang, C.-L.; Gu, L.; and Han, F. 2023. A simple and efficient pipeline to build an end-to-end spatial-temporal action detector. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 5999--6008

    2023

  40. [48]

    Tang, Y.; Chen, W.; Luo, Y.; and Zhang, Y. 2021. Humble Teachers Teach Better Students for Semi-Supervised Object Detection. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 3131--3140

    2021

  41. [49]

    Tarvainen, A.; and Valpola, H. 2017. Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results. In NIPS

    2017

  42. [50]

    Wang, X.; Zhang, S.; Qing, Z.; Shao, Y.; Gao, C.; and Sang, N. 2021. Self-Supervised Learning for Semi-Supervised Temporal Action Proposal. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 1905--1914

    2021

  43. [51]

    Weinzaepfel, P.; Harchaoui, Z.; and Schmid, C. 2015. Learning to track for spatio-temporal action localization. In Proceedings of the IEEE international conference on computer vision, 3164--3172

    2015

  44. [52]

    Wu, T.; Cao, M.; Gao, Z.; Wu, G.; and Wang, L. 2023. STMixer: A One-Stage Sparse Action Detector. ArXiv, abs/2303.15879

    2023 arXiv

  45. [53]

    Xiao, J.; Jing, L.; Zhang, L.; He, J.; She, Q.; Zhou, Z.; Yuille, A.; and Li, Y. 2022. Learning From Temporal Gradient for Semi-Supervised Action Recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 3252--3262

    2022

  46. [54]

    Xu, M.; Zhang, Z.; Hu, H.; Wang, J.; Wang, L.; Wei, F.; Bai, X.; and Liu, Z. 2021. End-to-End Semi-Supervised Object Detection with Soft Teacher. ArXiv, abs/2106.09018

    2021 arXiv

  47. [55]

    Xu, N.; Yang, L.; Fan, Y.; Yang, J.; Yue, D.; Liang, Y.; Price, B.; Cohen, S.; and Huang, T. 2018 a . Youtube-vos: Sequence-to-sequence video object segmentation. In Proceedings of the European Conference on Computer Vision (ECCV), 585--601

    2018

  48. [56]

    L.; Cohen, S

    Xu, N.; Yang, L.; Fan, Y.; Yang, J.; Yue, D.; Liang, Y.; Price, B. L.; Cohen, S. D.; and Huang, T. S. 2018 b . YouTube-VOS: Sequence-to-Sequence Video Object Segmentation. ArXiv, abs/1809.00461

    2018 arXiv

  49. [57]

    Xu, N.; Yang, L.; Fan, Y.; Yue, D.; Liang, Y.; Yang, J.; and Huang, T. S. 2018 c . YouTube-VOS: A Large-Scale Video Object Segmentation Benchmark. ArXiv, abs/1809.03327

    2018 arXiv

  50. [58]

    Xu, Y.; Wei, F.; Sun, X.; Yang, C.; Shen, Y.; Dai, B.; Zhou, B.; and Lin, S. 2022. Cross-Model Pseudo-Labeling for Semi-Supervised Action Recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2959--2968

    2022

  51. [59]

    Yang, J.; and Dai, K. 2023. Yowov2: A stronger yet efficient multi-level detection framework for real-time spatio-temporal action detection. arXiv preprint arXiv:2302.06848

    2023 arXiv

  52. [60]

    S.; and Kautz, J

    Yang, X.; Yang, X.; Liu, M.-Y.; Xiao, F.; Davis, L. S.; and Kautz, J. 2019. Step: Spatio-temporal progressive learning for video action detection. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 264--272

    2019

  53. [61]

    Yang, Z.; Gao, J.; and Nevatia, R. 2017. Spatio-temporal action detection with cascade proposal and location anticipation. In Proceedings of the British Machine Vision Conference ( BMVC )

    2017

  54. [62]

    Zhang, S.; Song, L.; Gao, C.; and Sang, N. 2020. GLNet: Global Local Network for Weakly Supervised Action Localization. IEEE Transactions on Multimedia, 22(10): 2610--2622

    2020

  55. [63]

    G.; and Tighe, J

    Zhao, J.; Li, X.; Liu, C.; Bing, S.; Chen, H.; Snoek, C. G.; and Tighe, J. 2021. Tuber: Tube-transformer for action detection. arXiv preprint arXiv:2104.00969

    2021 arXiv

  56. [64]

    Zhao, J.; Zhang, Y.; Li, X.; Chen, H.; Shuai, B.; Xu, M.; Liu, C.; Kundu, K.; Xiong, Y.; Modolo, D.; Marsic, I.; Snoek, C. G. M.; and Tighe, J. 2022. TubeR: Tubelet Transformer for Video Action Detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern...

    2022

  57. [65]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all...

  58. [66]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.