Pith. sign in

REVIEW 4 major objections 6 minor 3 cited by

Rethinking Pseudo-Label Guided Learning for Weakly Supervised Temporal Action Localization from the Perspective of Noise Correction

T0 review · 4 major / 6 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 NoCo corrects three structural types of pseudo-label noise—inaccurate boundaries, missing short clips, and merged adjacent instances—and sets new state-of-the-art results for weakly supervised temporal action localization, while running…

desk verdict A solid, well-ablated WTAL method that reframes pseudo-label learning as noise correction, but the paper overclaims the mechanism without reporting pseudo-label quality metrics. read the letter →

arxiv 2501.11124 v2 pith:Z63ZQWLV submitted 2025-01-19 cs.CV

Quan Zhang , Yuxin Qi , Xi Tang , Rui Yuan , Xi Lin , Ke Zhang , Chun Yuan This is my paper · ORCID
classification cs.CV
keywords weaklysupervisedtemporalactionlocalizationpseudo-labellearningnoisecorrectionteacher-studentframeworkcontext-awarelabelaugmentationmissinginstancecompensationambiguousnoisy
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

Pseudo-labels generated by weakly supervised temporal action localization (WTAL) models are noisy in three systematic ways: boundaries are imprecise, short action clips are missed, and adjacent instances get merged into one. The paper argues that this noise—not the weakly supervised model itself—is the main bottleneck when such pseudo-labels are used to train a fully supervised detection head, and that each noise type can be corrected without any extra supervision. To that end it proposes NoCo, a two-stage pipeline: a context-aware label augmentation step that sharpens boundaries using neighboring proposals, followed by an online teacher–student framework whose teacher (an EMA of past students) compensates missing instances, resolves merged instances, and reweights labels by quality. On THUMOS14 and ActivityNet v1.2 the paper reports average mAP of 50.9% and 30.7%, respectively, surpassing prior WTAL state-of-the-art methods, with a roughly four-fold inference speedup because only the lightweight student runs at test time.

What carries the argument

The load-bearing mechanism is the online teacher–student correction loop, in which the teacher is an exponential moving average of historical student checkpoints, giving it stable confidence and IoU signals from which to correct the current pseudo-label set. Within that loop, two modules consume teacher predictions: the Ambiguous Instance Correction module (AIC) forms, for each existing pseudo-label, an associated set of teacher proposals sharing the same category with IoU above a threshold, then fuses them via weighted averaging of start and end timestamps; the Missing Instance Compensation module (MIC) adds teacher proposals whose maximum IoU against all existing pseudo-labels falls below a threshold, with an adaptive confidence-based weight. The High-quality Pseudo-label Mining loss (HPM) then scales the classification and regression loss of each positive proposal by a weight derived from its IoU and confidence, so low-quality labels contribute less. Before the loop, the Context-Aware Label Augmentation module (CALA) replaces plain NMS with a weighted aggregation of neighboring same-class proposals sharing IoU above a threshold, using the weight $w_r = e^{\sqrt{\mathrm{IoU}} \cdot \min(\max(\theta_r,0),1)}$ and separate fusion functions for start and end timestamps, which the paper claims recovers boundary context that NMS discards.

What would settle it

On THUMOS14, where ground-truth annotations are available, compare per-instance IoU with ground truth for (a) the raw pseudo-labels, (b) the teacher's online predictions, and (c) the corrected labels after AIC and MIC: if the teacher's predictions are not systematically closer to ground truth than the pseudo-labels they correct—especially on the low-IoU instances MIC adds and the overlapping instances AIC adjusts—then the correction signal is not actually more trustworthy, and the reported gains would not be attributable to noise correction.

Watch

Extended reading notes

Core claim

The central discovery is that pseudo-label noise in WTAL is not random but consists of three identifiable structural errors—inaccurate instance boundaries, missing short action clips, and adjacent instances collapsed into one—and each can be corrected by a dedicated module that uses teacher predictions as a second opinion. The Ambiguous Instance Correction module mines teacher proposals with high IoU against an existing pseudo-label and re-aggregates their start and end times to tighten boundaries; the Missing Instance Compensation module adds teacher proposals whose IoU with all existing pseudo-labels is low, recovering missed clips; and the High-quality Pseudo-label Mining loss weights each positive proposal by a quality score so clean labels dominate optimization. The paper demonstrates this on two benchmarks: applying NoCo to the ASM-Loc baseline raises average mAP from 45.1% to 50.9% on THUMOS14 and from 26.5% to 30.7% on ActivityNet v1.2, surpassing the previous state-of-the-art on both benchmarks, and the same wrapper improves other WTAL bases (UM, CO2-Net) and point-supervised methods (LACP, HR-Pro, BackTAL). Because the noise-correction modules sit outside the WTAL base model, inference needs only the TriDet student, giving a 6.67 video/s speed versus 1.64 video/s for the baseline.

Load-bearing premise

The correction loop assumes that a teacher model trained on the very same noisy pseudo-labels it is asked to fix produces confidence and IoU signals that are more trustworthy than those labels; if the teacher's errors correlate with the pseudo-label noise, the corrections could amplify the original mistakes rather than remove them.

Editorial extensions

If this is right

  • NoCo sets a new state of the art on THUMOS14 and ActivityNet v1.2 for weakly supervised temporal action localization, with average mAP of 50.9% and 30.7%.
  • Because the noise-correction wrapper is decoupled from the WTAL base model, it can be grafted onto other weakly supervised methods; the paper shows consistent gains for UM, CO2-Net, ASM-Loc, and the point-supervised methods LACP, HR-Pro, and BackTAL.
  • Inference requires only the lightweight TriDet student, so NoCo runs at 6.67 video/s versus 1.64 video/s for the baseline WTAL method, with FLOPS reduced from 124.79G to 20.12G.
  • Each noise type has a dedicated module, and the ablations attribute specific gains to each: CALA adds 1.3% average mAP, the teacher–student scheme 0.7%, AIC 2.2%, MIC 1.8%, and HPM 0.2% on THUMOS14.
  • On ActivityNet v1.2 NoCo also outperforms some fully supervised methods such as SSN, suggesting that correcting pseudo-label noise can narrow part of the gap between weakly and fully supervised temporal action localization.

Reading between the lines

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

  • The three noise types are generic to any weakly supervised proposal-generation pipeline, so the AIC/MIC/HPM recipe could plausibly transfer to weakly supervised object detection in video or spatio-temporal action localization; the paper does not test this.
  • Because the teacher aggregates historical student checkpoints, correction quality likely depends on how quickly the student's errors decorrelate across iterations, so the EMA update rate is a critical hyperparameter that may need per-dataset tuning beyond the settings reported.
  • The HPM loss reweights only positive foreground proposals; a natural untested extension would apply the same quality signal to background proposals, which might reduce false positives from context-heavy scenes.
  • If the teacher's predictions remain more accurate than pseudo-labels on out-of-distribution videos, NoCo could serve as a label-correction wrapper for deploying WTAL to new domains with only video-level labels, but the reported experiments are limited to two standard benchmarks.
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 / 6 minor

Summary. The paper proposes NoCo, a weakly supervised temporal action localization method built on pseudo-label noise correction. It first generates instance-level pseudo-labels with a WTAL base model and refines their boundaries with a Context-Aware Label Augmentation (CALA) module, which aggregates neighboring same-class proposals using IoU- and confidence-based weights. It then trains a fully supervised student detector under an online teacher-student framework, where the teacher is an EMA of historical students and provides correction signals through two modules: Ambiguous Instance Correction (AIC), which refines existing pseudo-labels by aggregating high-IoU teacher predictions, and Missing Instance Compensation (MIC), which adds low-IoU teacher proposals as new instances. A High-quality Pseudo-label Mining (HPM) loss reweights positive samples during training. The method is evaluated on THUMOS14 and ActivityNet v1.2, reporting state-of-the-art average mAP of 50.9% and 30.7% respectively, and is shown to generalize when plugged into ASM-Loc, UM, CO2-Net, and two point-supervised baselines. The paper also claims a roughly threefold inference-speed improvement over its baseline.

Significance. If the results hold, the paper addresses a real and under-studied problem in pseudo-label-based WTAL: the noise in automatically generated instance labels. The design is modular and decoupled from the WTAL base model, and the generalization experiments across four base models strengthen the practical relevance. The two-stage pipeline with CALA, AIC, MIC, and HPM is coherent, and the ablations are internally consistent with the claimed contributions. The paper also reports a useful practical benefit: inference uses only the lightweight student detector. The main weaknesses are evidentiary: the central claim that gains come from noise correction is not directly supported by pseudo-label quality measurements, the speed comparison is not controlled across architectures, and the method's many thresholds are not subjected to sensitivity analysis or variance reporting. These issues leave the quantitative claims plausible but not fully established.

major comments (4)
  1. [Implementation; Online Noise Correction Framework] The Implementation section states that the authors 'report mIoU for pseudo-label quality,' but no such numbers appear in the main text or tables. The paper's central claim is that CALA, AIC, and MIC correct the three identified types of pseudo-label noise. Without direct measurements showing that the corrected pseudo-labels are closer to ground truth (e.g., mIoU before CALA, after CALA, and after AIC/MIC, and separately for boundary error, missing instances, and many-to-one cases), the attribution of the reported mAP gains to noise correction is not established. This is especially important because the teacher model is pretrained on the same CALA-enhanced pseudo-labels that AIC and MIC are supposed to correct; if the teacher inherits the same error patterns, correction could mainly average or duplicate existing noise. Please provide pseudo-label quality metrics and, if possible, a comparison between teacher prediction quality and initial pseudo-label quality.
  2. [Table 5; Analysis of Inference Speed] The speed comparison is not apples-to-apples. Table 5 compares a 'Baseline' at 35.8 mAP, 1.64 video/s, and 124.79 GFLOPs with 'NoCo' at 42.1 mAP, 6.67 video/s, and 20.12 GFLOPs. The baseline is the ASM-Loc-based pseudo-label retraining pipeline, while NoCo uses the lightweight TriDet student at inference. The 'threefold improvement' therefore conflates the effect of the proposed modules with the choice of a different, more efficient detector architecture. To support the speed claim, report inference speed of the same student detector with and without NoCo's correction modules, and also report speed for prior state-of-the-art methods under the same hardware and feature-extraction setup.
  3. [Experiments; Ablation Studies] The method introduces a large number of free hyperparameters: confidence threshold psi, NMS threshold rho, IoU thresholds eta0, eta1, and eta2, blending weight alpha, exponent beta, and positive loss weight lambda. No sensitivity analysis is reported for any of these, and no standard deviations over multiple seeds are provided for the main results or the ablation rows. Given that the state-of-the-art margins are small (e.g., 1.2 mAP over PivoTAL on THUMOS14 and 1.1 mAP over Ju et al. on ActivityNet v1.2), it is important to show that the conclusions are robust to threshold choices and random initialization. Please report at least three-seed mean and standard deviation for the main tables and a sensitivity study for the most critical thresholds (eta1, eta2, alpha).
  4. [Methods; Online Noise Correction Framework] The main text repeatedly defers algorithmic details to the supplementary material: 'We refer to Supplementary for the workflow of AIC' and 'We refer to Supplementary for the main workflow of the MIC.' In the provided manuscript, no supplementary file is present. Equations (4)-(7) give the aggregation and weighting formulas, but the full algorithms, the order in which AIC and MIC update the online pseudo-label set, and the handling of weights across iterations are not completely specified. This is a reproducibility issue for a central component of the method; the workflows should be included in the main text or the supplementary material should be provided.
minor comments (6)
  1. [Methods; Framework Overview] There is a duplicated word in the introduction: 'perceives the contextual information for each pseudo-label and and corrects them' should read 'and corrects them.'
  2. [Conclusion] The section heading 'Conslusion' is a typo and should be 'Conclusion.'
  3. [Table 3; Ablation Studies] The 'Baseline' row in Table 3 deserves a clearer label. The value 35.8 is lower than ASM-Loc's own reported 45.1 mAP in Table 1, which may confuse readers. Please clarify that this row is the TriDet pseudo-label retraining baseline without any correction modules and explain why this pipeline underperforms the original WTAL model.
  4. [Results on THUMOS14] The text says NoCo 'surpasses the latest methods CASE, AHIM, DDG-Net, and the current leading method PivoTAL,' but neither AHIM nor DDG-Net appears in Table 1 or in the reference list. Either add the corresponding rows and citations or remove the names.
  5. [Table 1; Table 2] The column header 'A VG' appears to be a formatting artifact for 'Avg.' The caption should spell out that this is the average mAP over the listed IoU thresholds, as the text in Section 5 does.
  6. [Abstract; Results on THUMOS14] The abstract states that NoCo 'outperforms the previous state-of-the-art method in detection accuracy and inference speed greatly.' The accuracy margin over PivoTAL on THUMOS14 is 1.2 mAP, and the speed comparison is not controlled (see major comment on Table 5). Please temper the claim or provide additional supporting comparisons.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: NoCo is a self-training pipeline evaluated against external benchmarks, and no predicted quantity reduces to a fitted input or a self-citation chain.

full rationale

The derivation chain in NoCo is not circular. The pseudo-labels are generated by an externally trained WTAL base model (ASM-Loc), refined by CALA using the same pseudo-label set, and then used to train a teacher-student framework whose student is evaluated on held-out test data (THUMOS14 test set and ActivityNet v1.2 validation set). The final reported quantities are mAP values computed against ground-truth annotations, not quantities fitted or defined by the method's own equations. Although the teacher model is pretrained on the CALA-augmented pseudo-labels and then used by AIC and MIC to correct those very labels, this is a self-training refinement loop, not a definitional reduction: the teacher's predictions are a learned function of video features, not identical by construction to the input pseudo-labels. Equations (4) and (6) show interpolation between existing pseudo-labels and teacher predictions, but the teacher is not the identity map, so the corrected labels are not equal to the inputs by definition. No load-bearing self-citation appears: the cited prior works by the authors (Zhang and Qi 2024; Zhang et al. 2024b) are used only as general references for WTAL and deep-learning applications, not to justify the core noise-correction claim. The paper does state in the Implementation section that it will 'report mIoU for pseudo-label quality,' yet no such numbers appear in the main text; this is a missing-support issue for attributing the gains to noise correction, but it is an empirical validation gap rather than a circularity. Because the method is benchmarked externally and no theoretical quantity is derived from itself, the appropriate circularity score is 0.

Assumptions & free parameters 8 free parameters · 4 assumptions · 0 invented entities

The framework introduces no new physical entities; everything the central claim rests on is a set of manual thresholds, an assumption about pseudo-label error structure, and a self-training loop. Hyperparameters for CALA, AIC, MIC, and HPM are listed but their values are not given in the main text, and no sensitivity analysis is reported.

free parameters (8)
  • Confidence threshold psi (CALA) = not reported in main text
    Filters pseudo-label instances before NMS in Algorithm 1; value deferred to supplementary.
  • NMS threshold rho (CALA) = not reported
    Suppresses overlapping pseudo-label instances; affects which instances are kept.
  • IoU threshold eta0 (CALA association) = not reported
    Defines the associated action instance set in CALA.
  • IoU threshold eta1 (AIC association) = not reported
    Controls which teacher predictions are fused into each online pseudo-label.
  • IoU threshold eta2 (MIC compensation) = not reported
    Determines whether a teacher instance is considered missing.
  • Blending weight alpha (AIC) = not reported
    Equation 4 mixes original pseudo-label boundary with teacher-associated context.
  • Exponent beta (AIC weight) = not reported
    Equation 6 controls corrected weight magnitude.
  • Positive loss weight lambda (HPM) = not reported
    Weights positive versus background classification in Equation 8.
assumptions (4)
  • domain assumption I3D features pretrained on Kinetics400 are a sufficient fixed representation for action localization.
    Used throughout without fine-tuning or comparison to other features.
  • domain assumption A WTAL base model's pseudo-labels contain exactly the three noise types targeted, and correcting these types is sufficient.
    The whole framework is built around boundary, missing, and many-to-one noise; no quantitative noise analysis is provided.
  • domain assumption Teacher predictions derived from the same noisy pseudo-label pipeline are more reliable than the pseudo-labels themselves.
    AIC and MIC trust teacher outputs for correcting and completing the online label set.
  • ad hoc to paper Pseudo-label thresholds and weighting formulas generalize across classes and datasets without re-tuning.
    All thresholds are fixed by hand and no sensitivity analysis is reported.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Rethinking Pseudo-Label Guided Learning for Weakly Supervised Temporal Action Localization from the Perspective of Noise Correction." pith.science (2026). https://pith.science/paper/Z63ZQWLV

@misc{pith2026250111124,
  author       = {Pith},
  title        = {Pith review of: Rethinking Pseudo-Label Guided Learning for Weakly Supervised Temporal Action Localization from the Perspective of Noise Correction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Z63ZQWLV}},
  note         = {Machine review of arXiv:2501.11124}
}
read the original abstract

Pseudo-label learning methods have been widely applied in weakly-supervised temporal action localization. Existing works directly utilize weakly-supervised base model to generate instance-level pseudo-labels for training the fully-supervised detection head. We argue that the noise in pseudo-labels would interfere with the learning of fully-supervised detection head, leading to significant performance leakage. Issues with noisy labels include:(1) inaccurate boundary localization; (2) undetected short action clips; (3) multiple adjacent segments incorrectly detected as one segment. To target these issues, we introduce a two-stage noisy label learning strategy to harness every potential useful signal in noisy labels. First, we propose a frame-level pseudo-label generation model with a context-aware denoising algorithm to refine the boundaries. Second, we introduce an online-revised teacher-student framework with a missing instance compensation module and an ambiguous instance correction module to solve the short-action-missing and many-to-one problems. Besides, we apply a high-quality pseudo-label mining loss in our online-revised teacher-student framework to add different weights to the noisy labels to train more effectively. Our model outperforms the previous state-of-the-art method in detection accuracy and inference speed greatly upon the THUMOS14 and ActivityNet v1.2 benchmarks.

Figures

Figures reproduced from arXiv: 2501.11124 by the authors.

Figure 1
Figure 1. Noise correction modeling for WTAL. NoCo in [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. (a). Framework Overview. The dark gray modules indicate the base method (eg. WTAL model and augmentation), [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Visualization of ground-truth and predictions. [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 3 Pith papers

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

  1. ComRoPE: Scalable and Robust Rotary Position Embedding Parameterized by Trainable Commuting Angle Matrices

    cs.CV 2025-06 conditional novelty 6.0 of 10

    Trainable commuting angle matrices generalize Rotary Position Embedding and improve accuracy and resolution robustness on vision Transformers.

  2. IMDPrompter: Adapting SAM to Image Manipulation Detection by Cross-View Automated Prompt Learning

    cs.CV 2025-02 conditional novelty 6.0 of 10

    IMDPrompter learns cross-view prompts for SAM from RGB, SRM, Bayer, and Noiseprint features, and reports state-of-the-art image manipulation detection and localization on five benchmarks.

  3. Modeling Multi-Task Model Merging as Adaptive Projective Gradient Descent

    cs.LG 2025-01 conditional novelty 6.0 of 10

    DOGE merges fine-tuned models by optimizing a data-free loss-gap proxy with gradient steps projected orthogonal to a shared task subspace, improving average accuracy over previous methods.

Reference graph

Works this paper leans on

49 extracted references · 42 canonical work pages · cited by 3 Pith papers

  1. [1]

    Carreira, J.; and Zisserman, A. 2017. Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset. In 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

    work page 2017

  2. [2]

    A.; Deng, J.; and Sukthankar, R

    Chao, Y.-W.; Vijayanarasimhan, S.; Seybold, B.; Ross, D. A.; Deng, J.; and Sukthankar, R. 2018. Rethinking the faster r-cnn architecture for temporal action localization. In Proceedings of the IEEE conference on computer vision and pattern recognition, 1130--1139

    work page 2018

  3. [3]

    Chen, M.; Gao, J.; Yang, S.; and Xu, C. 2022. Dual-evidential learning for weakly-supervised temporal action localization. In European Conference on Computer Vision, 192--208. Springer

    work page 2022

  4. [4]

    Chen, Y.; Huang, X.; Zhang, Q.; Li, W.; Zhu, M.; Yan, Q.; Li, S.; Chen, H.; Hu, H.; Yang, J.; et al. 2024. GIM: A Million-scale Benchmark for Generative Image Manipulation Detection and Localization. arXiv preprint arXiv:2406.16531

    arXiv 2024

  5. [5]

    Das, D.; and Lee, C. G. 2018. Graph matching and pseudo-label guided deep unsupervised domain adaptation. In Artificial Neural Networks and Machine Learning--ICANN 2018: 27th International Conference on Artificial Neural Networks, Rhodes, Greece, October 4-7, 2018, Proceedings, Part III 27, 342--352. Springer

    work page 2018

  6. [6]

    He, B.; Yang, X.; Kang, L.; Cheng, Z.; Zhou, X.; and Shrivastava, A. 2022. Asm-loc: Action-aware segment modeling for weakly-supervised temporal action localization. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 13925--13935

    work page 2022

  7. [7]

    Hong, F.-T.; Feng, J.-C.; Xu, D.; Shan, Y.; and Zheng, W.-S. 2021. Cross-modal consensus network for weakly supervised temporal action localization. In Proceedings of the 29th ACM international conference on multimedia, 1591--1599

    work page 2021

  8. [8]

    Ju, C.; Zhao, P.; Chen, S.; Zhang, Y.; Wang, Y.; and Tian, Q. 2021. Divide and conquer for single-frame temporal action localization. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 13455--13464

    work page 2021

Show all 49 references
  1. [9]

    Ju, C.; Zheng, K.; Liu, J.; Zhao, P.; Zhang, Y.; Chang, J.; Tian, Q.; and Wang, Y. 2023. Distilling Vision-Language Pre-training to Collaborate with Weakly-Supervised Temporal Action Localization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogni...

    2023

  2. [10]

    Lee, P.; and Byun, H. 2021. Learning action completeness from points for weakly-supervised temporal action localization. In Proceedings of the IEEE/CVF international conference on computer vision, 13648--13657

    2021

  3. [11]

    Lee, P.; Wang, J.; Lu, Y.; and Byun, H. 2021. Weakly-supervised temporal action localization by uncertainty modeling. In Proceedings of the AAAI conference on artificial intelligence, volume 35, 1854--1862

    2021

  4. [12]

    Li, J.; Socher, R.; and Hoi, S. C. 2020. Dividemix: Learning with noisy labels as semi-supervised learning. arXiv preprint arXiv:2002.07394

    2020 arXiv

  5. [13]

    Li, P.; Cao, J.; and Ye, X. 2023. Prototype contrastive learning for point-supervised temporal action detection. Expert Systems with Applications, 213: 118965

    2023

  6. [14]

    Liu, D.; Jiang, T.; and Wang, Y. 2019. Completeness modeling and context separation for weakly supervised temporal action localization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 1298--1307

    2019

  7. [15]

    Liu, Q.; Wang, Z.; Rong, S.; Li, J.; and Zhang, Y. 2023. Revisiting Foreground and Background Separation in Weakly-supervised Temporal Action Localization: A Clustering-based Approach. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 10433--10443

    2023

  8. [16]

    Liu, S.; Niles-Weed, J.; Razavian, N.; and Fernandez-Granda, C. 2020. Early-learning regularization prevents memorization of noisy labels. Advances in neural information processing systems, 33: 20331--20342

    2020

  9. [17]

    Liu, Z.; Wang, L.; Tang, W.; Yuan, J.; Zheng, N.; and Hua, G. 2021 a . Weakly supervised temporal action localization through learning explicit subspaces for action and context. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 35, 2242--2250

    2021

  10. [18]

    Liu, Z.; Wang, L.; Zhang, Q.; Tang, W.; Yuan, J.; Zheng, N.; and Hua, G. 2021 b . Acsnet: Action-context separation network for weakly supervised temporal action localization. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 35, 2233--2241

    2021

  11. [19]

    Lu, F.; Dong, S.; Zhang, L.; Liu, B.; Lan, X.; Jiang, D.; and Yuan, C. 2024 a . Deep Homography Estimation for Visual Place Recognition. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, 10341--10349

    2024

  12. [20]

    Lu, F.; Lan, X.; Zhang, L.; Jiang, D.; Wang, Y.; and Yuan, C. 2024 b . CricaVPR: Cross-image Correlation-aware Representation Learning for Visual Place Recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 16772--16782

    2024

  13. [21]

    Lu, F.; Zhang, L.; Lan, X.; Dong, S.; Wang, Y.; and Yuan, C. 2024 c . Towards seamless adaptation of pre-trained models for visual place recognition. arXiv preprint arXiv:2402.14505

    2024 arXiv

  14. [22]

    Luo, Z.; Guillory, D.; Shi, B.; Ke, W.; Wan, F.; Darrell, T.; and Xu, H. 2020. Weakly-Supervised Action Localization with Expectation-Maximization Multi-Instance Learning. In European Conference on Computer Vision, 729--745

    2020

  15. [23]

    Ma, F.; Zhu, L.; Yang, Y.; Zha, S.; Kundu, G.; Feiszli, M.; and Shou, Z. 2020. Sf-net: Single-frame supervision for temporal action localization. In Computer Vision--ECCV 2020: 16th European Conference, Glasgow, UK, August 23--28, 2020, Proceedings, Part IV 16, 420--437. Springer

    2020

  16. [24]

    S.; Yang, M.-H.; and Shao, L

    Narayan, S.; Cholakkal, H.; Hayat, M.; Khan, F. S.; Yang, M.-H.; and Shao, L. 2021. D2-net: Weakly-supervised action localization via discriminative embeddings and denoised activations. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 13608--13617

    2021

  17. [25]

    Neubeck, A.; and Van Gool, L. 2006. Efficient non-maximum suppression. In 18th international conference on pattern recognition (ICPR'06), volume 3, 850--855. IEEE

    2006

  18. [26]

    Nguyen, P.; Liu, T.; Prasad, G.; and Han, B. 2018. Weakly supervised action localization by sparse temporal pooling network. In Proceedings of the IEEE conference on computer vision and pattern recognition, 6752--6761

    2018

  19. [27]

    Pardo, A.; Alwassel, H.; Caba, F.; Thabet, A.; and Ghanem, B. 2021. Refineloc: Iterative refinement for weakly-supervised action localization. In Proceedings of the IEEE/CVF winter conference on applications of computer vision, 3319--3328

    2021

  20. [28]

    Paul, S.; Roy, S.; and Roy-Chowdhury, A. K. 2018. W-talc: Weakly-supervised temporal activity localization and classification. In Proceedings of the European Conference on Computer Vision (ECCV), 563--579

    2018

  21. [29]

    Pleiss, G.; Zhang, T.; Elenberg, E.; and Weinberger, K. Q. 2020. Identifying mislabeled data using the area under the margin ranking. Advances in Neural Information Processing Systems, 33: 17044--17056

    2020

  22. [30]

    Ren, H.; Yang, W.; Zhang, T.; and Zhang, Y. 2023. Proposal-based multiple instance learning for weakly-supervised temporal action localization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2394--2404

    2023

  23. [31]

    J.; Schwing, A

    Ren, Z.; Yu, Z.; Yang, X.; Liu, M.-Y.; Lee, Y. J.; Schwing, A. G.; and Kautz, J. 2020. Instance-aware, context-focused, and memory-efficient weakly supervised object detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 10598--10607

    2020

  24. [32]

    N.; Mittal, G.; Yu, Y.; Hall, M.; Sajeev, S.; Shah, M.; and Chen, M

    Rizve, M. N.; Mittal, G.; Yu, Y.; Hall, M.; Sajeev, S.; Shah, M.; and Chen, M. 2023. PivoTAL: Prior-Driven Supervision for Weakly-Supervised Temporal Action Localization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 22992--23002

    2023

  25. [33]

    Shi, D.; Zhong, Y.; Cao, Q.; Ma, L.; Li, J.; and Tao, D. 2023. Tridet: Temporal action detection with relative boundary modeling. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 18857--18866

    2023

  26. [34]

    Wang, L.; Xiong, Y.; Lin, D.; and Van Gool, L. 2017. Untrimmednets for weakly supervised action recognition and detection. In Proceedings of the IEEE conference on Computer Vision and Pattern Recognition, 4325--4334

    2017

  27. [35]

    Weng, Z.; Yang, X.; Li, A.; Wu, Z.; and Jiang, Y.-G. 2022. Semi-supervised vision transformers. In European Conference on Computer Vision, 605--620. Springer

    2022

  28. [36]

    Yang, L.; Han, J.; Zhao, T.; Lin, T.; Zhang, D.; and Chen, J. 2021 a . Background-click supervision for temporal action localization. IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(12): 9814--9829

    2021

  29. [37]

    Yang, W.; Zhang, T.; Mao, Z.; Zhang, Y.; Tian, Q.; and Wu, F. 2021 b . Multi-scale structure-aware network for weakly supervised temporal action detection. IEEE Transactions on Image Processing, 30: 5848--5861

    2021

  30. [38]

    Yang, W.; Zhang, T.; Yu, X.; Qi, T.; Zhang, Y.; and Wu, F. 2021 c . Uncertainty guided collaborative training for weakly supervised temporal action detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 53--63

    2021

  31. [39]

    Yun, W.; Qi, M.; Wang, C.; and Ma, H. 2024. Weakly-Supervised Temporal Action Localization by Inferring Salient Snippet-Feature. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, 6908--6916

    2024

  32. [40]

    Zhai, Y.; Wang, L.; Tang, W.; Zhang, Q.; Yuan, J.; and Hua, G. 2020. Two-Stream Consensus Network for Weakly-Supervised Temporal Action Localization. In European Conference on Computer Vision, 37--54

    2020

  33. [41]

    Zhang, C.; Cao, M.; Yang, D.; Chen, J.; and Zou, Y. 2021. Cola: Weakly-supervised temporal action localization with snippet contrastive learning. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 16010--16019

    2021

  34. [42]

    Zhang, H.; Wang, X.; Xu, X.; Qing, Z.; Gao, C.; and Sang, N. 2024 a . Hr-pro: Point-supervised temporal action localization via hierarchical reliability propagation. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, 7115--7123

    2024

  35. [43]

    Zhang, Q.; Liu, X.; Li, W.; Chen, H.; Liu, J.; Hu, J.; Xiong, Z.; Yuan, C.; and Wang, Y. 2024 b . Distilling Semantic Priors from SAM to Efficient Image Restoration Models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 25409--25419

    2024

  36. [44]

    Zhang, Q.; and Qi, Y. 2024. Can MLLMs Guide Weakly-Supervised Temporal Action Localization Tasks? arXiv preprint arXiv:2411.08466

    2024 arXiv

  37. [45]

    Zhao, T.; Han, J.; Yang, L.; and Zhang, D. 2022. Equivalent classification mapping for weakly supervised temporal action localization. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(3): 3019--3031

    2022

  38. [46]

    Zhao, Y.; Xiong, Y.; Wang, L.; Wu, Z.; Tang, X.; and Lin, D. 2017. Temporal action detection with structured segment networks. In Proceedings of the IEEE international conference on computer vision, 2914--2923

    2017

  39. [47]

    Zhou, J.; Huang, L.; Wang, L.; Liu, S.; and Li, H. 2023. Improving Weakly Supervised Temporal Action Localization by Bridging Train-Test Gap in Pseudo Labels. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 23003--23012

    2023

  40. [48]

    , " * 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...

  41. [49]

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