Pith. sign in

REVIEW 3 major objections 6 minor 1 cited by

Context-aware TFL: A Universal Context-aware Contrastive Learning Framework for Temporal Forgery Localization

T0 review · 3 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read The paper argues that temporal forgery localization is best treated as intra-sample anomaly detection: a forged instant is one whose features stray from the majority of instants, and it builds a contrastive learning framework around that…

desk verdict A plausible and well-ablated contrastive framework for temporal forgery localization, with a solid average-SOTA claim that is undercut by a table contradiction and an unspecified baseline protocol. read the letter →

arxiv 2506.08493 v1 pith:RZTVUW6H submitted 2025-06-10 cs.CV cs.MM

Qilin Yin , Wei Lu , Xiangyang Luo , Xiaochun Cao This is my paper
classification cs.CVcs.MM
keywords temporalforgerylocalizationcontrastivelearninganomalydetectionaudio-visualforensicsmultimediapartialdeepfakeaction
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's central claim is that temporal forgery localization should be treated as intra-sample anomaly detection: in a partially tampered video, a forged temporal instant is one whose features stand far from the majority of instants, i.e., from the global context. Building on that premise, the paper proposes UniCaCLF, a contrastive learning framework with a context-aware perception layer and a context-aware contrastive loss, and asserts it outperforms prior TFL and temporal action localization methods on five benchmarks. The largest reported gains appear at the strictest evaluation point, AP@IoU=0.95, where the method reaches 53.61% versus 37.22% for UMMAFormer on LAV-DF, and 52.71% versus 41.17% on AV-Deepfake1M. If correct, this shows that contrasting each instant against a carefully re-estimated global context yields more discriminative features for locating small forged segments than reconstruction- or attention-based approaches.

What carries the argument

The load-bearing object is the context-aware perception (CaP) layer together with the context-aware contrastive loss (CaCL). The CaP layer's heterogeneous activation operation (HAO) computes the negative cosine similarity between each instant feature and the global context, applies ReLU to keep only the deviating instants, and re-weights those features so forged instants become more salient; the adaptive context updater (ACU) then forms a cleaned global context by softmax-weighting instants by their similarity to the previous context. CaCL uses that cleaned context as the query in a supervised contrastive objective computed sample-by-sample, pulling genuine instants closer and forged instants farther. These components run inside a multi-scale feature pyramid whose outputs feed classification and boundary-regression heads.

What would settle it

Take a set of partially tampered videos in which the forged clips are known to be statistically undistinguishable from genuine clips in TSN, BYOL-A, ResNet50, and Wave2vec features (for example, forged-to-genuine cosine similarity to the global context is equal in distribution). If UniCaCLF's localization accuracy then falls to the level of random guessing while a reconstruction-based baseline still localizes them, the central outlier-distance claim is refuted.

Watch

Extended reading notes

Core claim

UniCaCLF discovers forged temporal segments by treating them as outliers relative to the global context of the same input sample, not by comparing across samples. The context-aware perception layer strengthens instant features whose negative cosine similarity to the global context is positive, while an adaptive context updater recomputes the context so forged instants contaminate it as little as possible. A sample-by-sample supervised contrastive loss then pulls genuine instants toward the global context and pushes forged instants away from it. The paper reports that this combination outperforms state-of-the-art competitors across visual-only, audio-only, and audio-visual partial forgery datasets, with particularly large advantages at the tight IoU=0.95 threshold.

Load-bearing premise

The framework assumes that a forged temporal segment always behaves as a statistical outlier in the pre-trained feature space, standing noticeably farther from the video's average features than genuine moments do; if tampering is seamless enough to keep forged features inside the normal distribution, the method has no signal to detect it.

Editorial extensions

If this is right

  • At the strictest overlap threshold (AP@IoU=0.95), UniCaCLF beats UMMAFormer by 16.39 points on LAV-DF and 11.54 points on AV-Deepfake1M, and beats TriDet by 6.01 points on the HAD audio-only dataset.
  • The same architecture and training recipe work for video-only, audio-only, and audio-visual partial forgery by swapping frozen pre-trained encoders, and it generalizes across datasets (HAD to Psynd) better than the compared baselines.
  • Because the contrastive loss is computed sample-by-sample, performance does not depend on forged segments from different manipulation types being comparable to each other within a batch.
  • The CaP layer avoids the self-attention rank-loss problem by replacing dense attention with similarity-based activation, which the paper reports makes the model faster and lighter than transformer-based TFL baselines.

Reading between the lines

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

  • A testable extension is to apply the same global-context contrast to other sequential forgery domains, such as manipulated text or sensor streams, whenever a frozen encoder produces embeddings in which tampered tokens are outliers.
  • The components' value could be isolated further by ablating the adaptive context updater against an oracle context computed from ground-truth genuine instants only; this would show how much remaining error comes from context contamination.
  • The single-global-context assumption may weaken when forged segments occupy a large fraction of the video, since the 'majority' becomes ambiguous; testing on videos with fake-to-real ratios approaching 1:1 would probe that boundary.
  • Pairing the outlier-based signal with a reconstruction-based consistency signal could cover forgeries that are semantically consistent but not statistically anomalous.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 6 minor

Summary. The paper proposes UniCaCLF, a universal context-aware contrastive learning framework for temporal forgery localization (TFL). The framework constructs a context-aware perception (CaP) layer with a heterogeneous activation operation (HAO) and an adaptive context updater (ACU), and adds a supervised context-aware contrastive loss (CaCL) computed sample-by-sample. Experiments are conducted on four benchmark datasets (LAV-DF, AV-Deepfake1M, TVIL, HAD) plus a cross-dataset generalization test on Psynd, reporting AP and AR metrics as well as runtime/parameter counts. The authors claim that UniCaCLF significantly outperforms state-of-the-art TFL and temporal action localization methods, with particularly large gains at AP@IoU=0.95.

Significance. If the reported results are robust, the paper makes a solid empirical contribution: it demonstrates that an anomaly-detection formulation of TFL, combined with intra-sample contrastive learning, yields strong localization accuracy while remaining computationally light (7.32M parameters, 4.56 GFLOPs, 35.08 FPS on TVIL). The ablation study on TVIL shows positive contributions from HAO, ACU, and CaCL, and the HAD-to-Psynd cross-dataset experiment is a useful generalization check. The central average-performance claim is supported by Tables I-VI. However, the current text contains a factual overstatement about the LAV-DF results, and the comparison protocol is not fully controlled, which weakens the attribution of the reported gains to the proposed components. These issues are fixable but need to be addressed before publication.

major comments (3)
  1. [Section IV-B.1, Table I] The text states that UniCaCLF "can outperform all compared methods under both AP and AR" on LAV-DF, but Table I contradicts this. At AP@IoU=0.5, UniCaCLF scores 97.81, below MFMS (98.47) and UMMAFormer (98.34); at AP@IoU=0.75, UniCaCLF scores 93.11, below MFMS (94.15) and UMMAFormer (93.54). The correct claim is that UniCaCLF achieves the best average AP and the best AR values, not that it wins every threshold. This overstatement should be corrected in the text and abstract.
  2. [Section IV-A.4 and Section III-E] The comparison protocol is not sufficiently controlled to attribute the reported SOTA gains to the proposed HAO/ACU/CaCL mechanism. The paper states only that "unified pre-trained features" were adopted for all models; it does not state whether the baselines were retrained with the same classification/regression losses, loss weights, optimizer schedules, feature-pyramid construction, or post-processing. In particular, UniCaCLF uses focal loss, DIoU regression loss, and Soft-NMS, while the baselines use their own heads and losses. Since AP@IoU=0.95 is highly sensitive to boundary precision, the large margins at AP@0.95 (e.g., 53.61 vs. 37.22 on LAV-DF) could stem from these protocol choices rather than from the proposed context-aware contrastive learning. Please provide matched-protocol comparisons, or ablate the DIoU loss and Soft-NMS on the baseline methods.
  3. [Section IV-D, Eq. (5), and Section IV-A.3] The temperature hyperparameter tau in the context-aware contrastive loss (Eq. 5) is never specified. Section IV-A.3 gives the number of CaP layers L=6, the loss weights phi_1=2 and phi_2=0.5, and the forged-segment labeling threshold, but tau is omitted. Because CaCL is a core contribution and its behavior depends strongly on tau, this omission prevents exact reproduction and makes it impossible to assess the sensitivity of the proposed method to this key parameter. Please report the value used in all experiments and, if possible, include a sensitivity study.
minor comments (6)
  1. [Table VII] Rows 3 and 4 have identical checkmark patterns (both show two checkmarks), so the reader cannot tell whether the third row is HAO+CaCL or HAO+ACU. This makes the marginal contributions of CaCL and ACU unreadable. Please relabel the rows unambiguously.
  2. [Table II] In the BA-TFD+ row, the AR@10 value appears as "6364" instead of "63.64" (missing decimal point).
  3. [Section IV-C.2 and Table VIII] The text says "the fewer the Cap layers, the better the model localization performance," but Table VIII shows that 6 and 7 layers give the best AP@0.95 and average AP, respectively. This statement should be revised to match the table.
  4. [Eq. (5) and Section III-D] The CaCL loss uses a dot product g·x/tau, while HAO uses negative cosine similarity. Please state explicitly whether the features are L2-normalized before the dot product; if they are, say so, since this affects the scale of the loss and the interpretation of tau.
  5. [Fig. 4] The caption and text say "we statistic the average cosine similarity" but do not describe how the SA, SGP, and CaP features were produced for this figure. Please specify the experimental setup so the figure can be interpreted and reproduced.
  6. [Section IV-A.4] The exclusion of DiMoDif and Vigo due to lack of released code is reasonable. However, please clarify whether the reported baseline numbers are taken from the original papers or from re-implementations using the unified pre-trained features; this is important for judging the fairness of the comparison.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: UniCaCLF's central claim is benchmark performance, evaluated independently of the training objective; the anomaly assumption is stated and enforced via supervised losses, with no derivation reducing to its inputs.

full rationale

The paper's load-bearing premise, that in content-driven partial forgery a forged instant tends to be farther from the global context than a genuine instant, is explicitly stated as an assumption in Sections I and III-B rather than derived from an equation. The model then actively enforces this property through HAO, ACU, and the supervised CaCL loss using ground-truth instant labels, so the claim is a training objective rather than a result smuggled in from its own premise. The central empirical claim is AP/AR performance on five public datasets, which is evaluated on test or validation data independently of the CaCL loss values; even the largest gains at AP@IoU=0.95 are computed from predicted intervals against external ground truth, not from the distances the loss optimizes. The ablation study and Fig. 4 inspect properties of the trained model and do not constitute a derivation of the main result. Self-citations [1], [3], and [7] appear only as related-work references and are not load-bearing for any argument. The skeptic's concern about uncontrolled baseline protocol (DIoU loss, Soft-NMS, unspecified temperature tau) is a correctness or reproducibility risk, not a circularity, and the paper does state that unified pre-trained features were adopted for all compared models. No step reduces, by the paper's own equations or by self-citation, to its own inputs.

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

The framework rests on the anomaly-distribution assumption rather than a mathematical derivation. The method's components are algorithmic and the reported gains are empirical. Free parameters are standard hyperparameters; the most consequential choice is the reliance on fixed pre-trained features and the assumption that forged instants are outliers relative to the sample's global context.

free parameters (5)
  • tau (temperature in CaCL) = not reported
    Temperature hyperparameter in the contrastive loss, Eq. (5); its value is never given in the implementation details.
  • phi_1 and phi_2 (loss weights) = 2 and 0.5
    Weights balancing classification, regression, and contrastive losses in Eq. (7); stated as defaults with no sensitivity analysis.
  • L (number of CaP layers) = 6
    Number of pyramid levels chosen via ablation on TVIL (Table VIII); the paper reports the framework is insensitive to this choice.
  • Forged-segment labeling threshold = 40%
    Instants are labeled forged if a segment contains more than 40% forged instants after downsampling labels; stated in Section IV-A.3.
  • beta (ACU trainable weight) = trainable
    Adaptive context updater uses a learned interpolation weight beta in Eq. (4); value is learned during training.
assumptions (4)
  • domain assumption In content-driven partial forgery, forged instant features are significantly more distant from the global context than genuine instant features.
    Core premise of the anomaly-detection design, stated in Sections I and III-B; HAO and CaCL are constructed to enforce this property in the learned representation.
  • domain assumption Pre-trained feature extractors (TSN, BYOL-A, ResNet50, Wave2vec) retain sufficient anomaly information for TFL on manipulated content.
    Features are frozen and used as input; if the extractors discard low-level tampering traces, the framework cannot recover them (Section IV-A.3).
  • domain assumption Each sample's tampering type has a distinct feature distribution, so cross-sample contrastive learning is harmful; this motivates sample-by-sample CaCL.
    Stated in Section III-D; support comes mainly from the CaCL ablation rather than a direct measurement of cross-sample interference.
  • standard math Standard optimization and architecture components (Adam, focal loss, DIoU, Soft-NMS) behave as expected in this setting.
    Borrowed from prior work and used without modification; no new mathematical guarantees are provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Context-aware TFL: A Universal Context-aware Contrastive Learning Framework for Temporal Forgery Localization." pith.science (2026). https://pith.science/paper/RZTVUW6H

@misc{pith2026250608493,
  author       = {Pith},
  title        = {Pith review of: Context-aware TFL: A Universal Context-aware Contrastive Learning Framework for Temporal Forgery Localization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RZTVUW6H}},
  note         = {Machine review of arXiv:2506.08493}
}
read the original abstract

Most research efforts in the multimedia forensics domain have focused on detecting forgery audio-visual content and reached sound achievements. However, these works only consider deepfake detection as a classification task and ignore the case where partial segments of the video are tampered with. Temporal forgery localization (TFL) of small fake audio-visual clips embedded in real videos is still challenging and more in line with realistic application scenarios. To resolve this issue, we propose a universal context-aware contrastive learning framework (UniCaCLF) for TFL. Our approach leverages supervised contrastive learning to discover and identify forged instants by means of anomaly detection, allowing for the precise localization of temporal forged segments. To this end, we propose a novel context-aware perception layer that utilizes a heterogeneous activation operation and an adaptive context updater to construct a context-aware contrastive objective, which enhances the discriminability of forged instant features by contrasting them with genuine instant features in terms of their distances to the global context. An efficient context-aware contrastive coding is introduced to further push the limit of instant feature distinguishability between genuine and forged instants in a supervised sample-by-sample manner, suppressing the cross-sample influence to improve temporal forgery localization performance. Extensive experimental results over five public datasets demonstrate that our proposed UniCaCLF significantly outperforms the state-of-the-art competing algorithms.

Figures

Figures reproduced from arXiv: 2506.08493 by the authors.

Figure 2
Figure 2. Diagrammatic overview of the proposed universal context-aware contrastive learning framework. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Illustration of the structure of HAO. ⊕ and ⊙ denote the element-wise addition and element-wise multiplication respectively [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 5
Figure 5. Visualization results of the LAV-DF and TVIL datasets. The start and end timestamps (in seconds) of the forged [PITH_FULL_IMAGE:figures/full_fig_p010_5.png] view at source ↗

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

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

  1. UniSkip-Mamba: A Frequency-Aware State Space Model for Audio-Visual Temporal Forgery Localization

    cs.CV 2026-07 conditional novelty 6.0 of 10

    Skip-scanning Mamba with unified audio-visual sequences reaches 63.4% AP@0.95 on LAV-DF and 63.58% mAP on AV-Deepfake1M by regularizing toward low/mid-frequency forgery cues.

Reference graph

Works this paper leans on

55 extracted references · 37 canonical work pages · cited by 1 Pith paper

  1. [1]

    Audio multi-view spoofing detection framework based on audio-text-emotion correlations,

    J. Wu, Q. Yin, Z. Sheng, W. Lu, J. Huang, and B. Li, “Audio multi-view spoofing detection framework based on audio-text-emotion correlations,” IEEE Transactions on Information Forensics and Security, 2024

    work page 2024

  2. [2]

    Learning from yourself: A self-distillation method for fake speech detection,

    J. Xue, C. Fan, J. Yi, C. Wang, Z. Wen, D. Zhang, and Z. Lv, “Learning from yourself: A self-distillation method for fake speech detection,” in ICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2023, pp. 1–5

    work page 2023

  3. [3]

    Dynamic difference learning with spatio-temporal correlation for deepfake video detection,

    Q. Yin, W. Lu, B. Li, and J. Huang, “Dynamic difference learning with spatio-temporal correlation for deepfake video detection,”IEEE Transactions on Information Forensics and Security, 2023

    work page 2023

  4. [4]

    Deepfake detection via inter-frame inconsistency recomposition and enhancement,

    C. Zhu, B. Zhang, Q. Yin, C. Yin, and W. Lu, “Deepfake detection via inter-frame inconsistency recomposition and enhancement,”Pattern Recognition, vol. 147, p. 110077, 2024

    work page 2024

  5. [5]

    Where deepfakes gaze at? spatial-temporal gaze inconsistency analysis for video face forgery detection,

    C. Peng, Z. Miao, D. Liu, N. Wang, R. Hu, and X. Gao, “Where deepfakes gaze at? spatial-temporal gaze inconsistency analysis for video face forgery detection,”IEEE Transactions on Information Forensics and Security, 2024. 11

    work page 2024

  6. [6]

    Semantic contextualization of face forgery: A new definition, dataset, and detection method,

    M. Zou, B. Yu, Y . Zhan, S. Lyu, and K. Ma, “Semantic contextualization of face forgery: A new definition, dataset, and detection method,”IEEE Transactions on Information Forensics and Security, 2025

    work page 2025

  7. [7]

    Fine- grained multimodal deepfake classification via heterogeneous graphs,

    Q. Yin, W. Lu, X. Cao, X. Luo, Y . Zhou, and J. Huang, “Fine- grained multimodal deepfake classification via heterogeneous graphs,” International Journal of Computer Vision, pp. 1–15, 2024

    work page 2024

  8. [8]

    Avoid-df: Audio-visual joint learning for detecting deepfake,

    W. Yang, X. Zhou, Z. Chen, B. Guo, Z. Ba, Z. Xia, X. Cao, and K. Ren, “Avoid-df: Audio-visual joint learning for detecting deepfake,”IEEE Transactions on Information Forensics and Security, vol. 18, pp. 2015– 2029, 2023

    work page 2015

Show all 55 references
  1. [9]

    Glitch in the matrix!

    Z. Cai, S. Ghosh, T. Gedeon, A. Dhall, K. Stefanov, and M. Hayat, “Glitch in the matrix!”: A large scale benchmark for content driven audio-visual forgery detection and localization,”arXiv preprint arXiv:2305.01979, 2023

    2023 arXiv

  2. [10]

    Not made for each other-audio-visual dissonance-based deepfake detection and local- ization,

    K. Chugh, P. Gupta, A. Dhall, and R. Subramanian, “Not made for each other-audio-visual dissonance-based deepfake detection and local- ization,” inProceedings of the 28th ACM international conference on multimedia, 2020, pp. 439–447

    2020

  3. [11]

    Audio-visual temporal forgery de- tection using embedding-level fusion and multi-dimensional contrastive loss,

    M. Liu, J. Wang, X. Qian, and H. Li, “Audio-visual temporal forgery de- tection using embedding-level fusion and multi-dimensional contrastive loss,”IEEE Transactions on Circuits and Systems for Video Technology, 2023

    2023

  4. [12]

    Dimodif: Discourse modality- information differentiation for audio-visual deepfake detection and lo- calization,

    C. Koutlis and S. Papadopoulos, “Dimodif: Discourse modality- information differentiation for audio-visual deepfake detection and lo- calization,”arXiv preprint arXiv:2411.10193, 2024

    2024 arXiv

  5. [13]

    Um- maformer: A universal multimodal-adaptive transformer framework for temporal forgery localization,

    R. Zhang, H. Wang, M. Du, H. Liu, Y . Zhou, and Q. Zeng, “Um- maformer: A universal multimodal-adaptive transformer framework for temporal forgery localization,” inProceedings of the 31st ACM Inter- national Conference on Multimedia, 2023, pp. 8749–8759

    2023

  6. [14]

    Mfms: Learning modality-fused and modality- specific features for deepfake detection and localization tasks,

    Y . Zhang, C. Miao, M. Luo, J. Li, W. Deng, W. Yao, Z. Li, B. Hu, W. Feng, T. Gonget al., “Mfms: Learning modality-fused and modality- specific features for deepfake detection and localization tasks,” in Proceedings of the 32nd ACM International Conference on Multimedia, 2024,...

    2024

  7. [15]

    Attention is not all you need: Pure attention loses rank doubly exponentially with depth,

    Y . Dong, J.-B. Cordonnier, and A. Loukas, “Attention is not all you need: Pure attention loses rank doubly exponentially with depth,” in International Conference on Machine Learning. PMLR, 2021, pp. 2793–2803

    2021

  8. [16]

    Av-deepfake1m: A large-scale llm-driven audio- visual deepfake dataset,

    Z. Cai, S. Ghosh, A. P. Adatia, M. Hayat, A. Dhall, T. Gedeon, and K. Stefanov, “Av-deepfake1m: A large-scale llm-driven audio- visual deepfake dataset,” inProceedings of the 32nd ACM International Conference on Multimedia, 2024, pp. 7414–7423

    2024

  9. [17]

    Half-truth: A partially fake audio detection dataset,

    J. Yi, Y . Bai, J. Tao, H. Ma, Z. Tian, C. Wang, T. Wang, and R. Fu, “Half-truth: A partially fake audio detection dataset,”arXiv preprint arXiv:2104.03617, 2021

    2021 arXiv

  10. [18]

    Localizing fake segments in speech,

    B. Zhang and T. Sim, “Localizing fake segments in speech,” in2022 26th International Conference on Pattern Recognition (ICPR). IEEE, 2022, pp. 3224–3230

    2022

  11. [19]

    Emotions don’t lie: An audio-visual deepfake detection method using affective cues,

    T. Mittal, U. Bhattacharya, R. Chandra, A. Bera, and D. Manocha, “Emotions don’t lie: An audio-visual deepfake detection method using affective cues,” inProceedings of the 28th ACM international conference on multimedia, 2020, pp. 2823–2832

    2020

  12. [20]

    Multimodal forgery detection using ensemble learning,

    A. Hashmi, S. A. Shahzad, W. Ahmad, C. W. Lin, Y . Tsao, and H.- M. Wang, “Multimodal forgery detection using ensemble learning,” in 2022 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC). IEEE, 2022, pp. 1524–1532

    2022

  13. [21]

    Using graph neural networks to improve generalization capability of the models for deepfake detection,

    H. She, Y . Hu, B. Liu, J. Li, and C.-T. Li, “Using graph neural networks to improve generalization capability of the models for deepfake detection,”IEEE Transactions on Information Forensics and Security, 2024

    2024

  14. [22]

    Joint audio-visual deepfake detection,

    Y . Zhou and S.-N. Lim, “Joint audio-visual deepfake detection,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 14 800–14 809

    2021

  15. [23]

    Avfakenet: A unified end-to-end dense swin transformer deep learning model for audio–visual deepfakes detection,

    H. Ilyas, A. Javed, and K. M. Malik, “Avfakenet: A unified end-to-end dense swin transformer deep learning model for audio–visual deepfakes detection,”Applied Soft Computing, vol. 136, p. 110124, 2023

    2023

  16. [24]

    Audio-visual person-of-interest deepfake detection,

    D. Cozzolino, A. Pianese, M. Nießner, and L. Verdoliva, “Audio-visual person-of-interest deepfake detection,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2023, pp. 943– 952

    2023

  17. [25]

    Contextloc++: A unified context model for temporal action localization,

    Z. Zhu, L. Wang, W. Tang, N. Zheng, and G. Hua, “Contextloc++: A unified context model for temporal action localization,”IEEE Transac- tions on Pattern Analysis and Machine Intelligence, vol. 45, no. 8, pp. 9504–9519, 2023

    2023

  18. [26]

    Deep learning-based action detection in untrimmed videos: A survey,

    E. Vahdani and Y . Tian, “Deep learning-based action detection in untrimmed videos: A survey,”IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022

    2022

  19. [27]

    Semantic and temporal contextual correlation learning for weakly-supervised temporal action localization,

    J. Fu, J. Gao, and C. Xu, “Semantic and temporal contextual correlation learning for weakly-supervised temporal action localization,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 45, no. 10, pp. 12 427–12 443, 2023

    2023

  20. [28]

    Gaussian temporal awareness networks for action localization,

    F. Long, T. Yao, Z. Qiu, X. Tian, J. Luo, and T. Mei, “Gaussian temporal awareness networks for action localization,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 344–353

    2019

  21. [29]

    Mid-level fusion for end-to-end temporal activity detection in untrimmed video

    M. A. Rahman and R. Lagani `ere, “Mid-level fusion for end-to-end temporal activity detection in untrimmed video.” inBMVC, 2020

    2020

  22. [30]

    R-c3d: Region convolutional 3d network for temporal activity detection,

    H. Xu, A. Das, and K. Saenko, “R-c3d: Region convolutional 3d network for temporal activity detection,” inProceedings of the IEEE international conference on computer vision, 2017, pp. 5783–5792

    2017

  23. [31]

    Rethinking the faster r-cnn architecture for temporal action localization,

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

    2018

  24. [32]

    Graph attention based proposal 3d convnets for action detection,

    J. Li, X. Liu, Z. Zong, W. Zhao, M. Zhang, and J. Song, “Graph attention based proposal 3d convnets for action detection,” inProceedings of the AAAI Conference on Artificial Intelligence, vol. 34, no. 04, 2020, pp. 4626–4633

    2020

  25. [33]

    Bsn: Boundary sensitive network for temporal action proposal generation,

    T. Lin, X. Zhao, H. Su, C. Wang, and M. Yang, “Bsn: Boundary sensitive network for temporal action proposal generation,” inProceedings of the European conference on computer vision (ECCV), 2018, pp. 3–19

    2018

  26. [34]

    Bmn: Boundary-matching network for temporal action proposal generation,

    T. Lin, X. Liu, X. Li, E. Ding, and S. Wen, “Bmn: Boundary-matching network for temporal action proposal generation,” inProceedings of the IEEE/CVF international conference on computer vision, 2019, pp. 3889–3898

    2019

  27. [35]

    Actionformer: Localizing moments of actions with transformers,

    C.-L. Zhang, J. Wu, and Y . Li, “Actionformer: Localizing moments of actions with transformers,” inEuropean Conference on Computer Vision. Springer, 2022, pp. 492–510

    2022

  28. [36]

    Tridet: Temporal action detection with relative boundary modeling,

    D. Shi, Y . Zhong, Q. Cao, L. Ma, J. Li, and D. Tao, “Tridet: Temporal action detection with relative boundary modeling,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 18 857–18 866

    2023

  29. [37]

    Temporal action localization in the deep learning era: A survey,

    B. Wang, Y . Zhao, L. Yang, T. Long, and X. Li, “Temporal action localization in the deep learning era: A survey,”IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023

    2023

  30. [38]

    A simple framework for contrastive learning of visual representations,

    T. Chen, S. Kornblith, M. Norouzi, and G. Hinton, “A simple framework for contrastive learning of visual representations,” inInternational conference on machine learning. PMLR, 2020, pp. 1597–1607

    2020

  31. [39]

    Exploring simple siamese representation learning,

    X. Chen and K. He, “Exploring simple siamese representation learning,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021, pp. 15 750–15 758

    2021

  32. [40]

    Momentum contrast for unsupervised visual representation learning,

    K. He, H. Fan, Y . Wu, S. Xie, and R. Girshick, “Momentum contrast for unsupervised visual representation learning,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 9729–9738

    2020

  33. [41]

    Representation learning with contrastive predictive coding,

    A. v. d. Oord, Y . Li, and O. Vinyals, “Representation learning with contrastive predictive coding,”arXiv preprint arXiv:1807.03748, 2018

    2018 arXiv

  34. [42]

    Unsupervised feature learning via non-parametric instance discrimination,

    Z. Wu, Y . Xiong, S. X. Yu, and D. Lin, “Unsupervised feature learning via non-parametric instance discrimination,” inProceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 3733– 3742

    2018

  35. [43]

    Fully unsupervised deepfake video detection via enhanced contrastive learning,

    T. Qiao, S. Xie, Y . Chen, F. Retraint, and X. Luo, “Fully unsupervised deepfake video detection via enhanced contrastive learning,”IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024

    2024

  36. [44]

    Frequency-aware discrim- inative feature learning supervised by single-center loss for face forgery detection,

    J. Li, H. Xie, J. Li, Z. Wang, and Y . Zhang, “Frequency-aware discrim- inative feature learning supervised by single-center loss for face forgery detection,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021, pp. 6458–6467

    2021

  37. [45]

    Dual contrastive learning for general face forgery detection,

    K. Sun, T. Yao, S. Chen, S. Ding, J. Li, and R. Ji, “Dual contrastive learning for general face forgery detection,” inProceedings of the AAAI Conference on Artificial Intelligence, vol. 36, no. 2, 2022, pp. 2316– 2324

    2022

  38. [46]

    Supervised contrastive learn- ing,

    P. Khosla, P. Teterwak, C. Wang, A. Sarna, Y . Tian, P. Isola, A. Maschinot, C. Liu, and D. Krishnan, “Supervised contrastive learn- ing,”Advances in neural information processing systems, vol. 33, pp. 18 661–18 673, 2020

    2020

  39. [47]

    Temporal segment networks: Towards good practices for deep action recognition,

    L. Wang, Y . Xiong, Z. Wang, Y . Qiao, D. Lin, X. Tang, and L. Van Gool, “Temporal segment networks: Towards good practices for deep action recognition,” inEuropean conference on computer vision. Springer, 2016, pp. 20–36

    2016

  40. [48]

    Quo vadis, action recognition? a new model and the kinetics dataset,

    J. Carreira and A. Zisserman, “Quo vadis, action recognition? a new model and the kinetics dataset,” inproceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017, pp. 6299–6308. 12

    2017

  41. [49]

    Byol for audio: Self-supervised learning for general-purpose audio represen- tation,

    D. Niizumi, D. Takeuchi, Y . Ohishi, N. Harada, and K. Kashino, “Byol for audio: Self-supervised learning for general-purpose audio represen- tation,” in2021 International Joint Conference on Neural Networks (IJCNN). IEEE, 2021, pp. 1–8

    2021

  42. [50]

    Focal loss for dense object detection,

    T.-Y . Lin, P. Goyal, R. Girshick, K. He, and P. Doll ´ar, “Focal loss for dense object detection,” inProceedings of the IEEE international conference on computer vision, 2017, pp. 2980–2988

    2017

  43. [51]

    Distance-iou loss: Faster and better learning for bounding box regression,

    Z. Zheng, P. Wang, W. Liu, J. Li, R. Ye, and D. Ren, “Distance-iou loss: Faster and better learning for bounding box regression,” inProceedings of the AAAI conference on artificial intelligence, vol. 34, no. 07, 2020, pp. 12 993–13 000

    2020

  44. [52]

    Soft-nms–improving object detection with one line of code,

    N. Bodla, B. Singh, R. Chellappa, and L. S. Davis, “Soft-nms–improving object detection with one line of code,” inProceedings of the IEEE international conference on computer vision, 2017, pp. 5561–5569

    2017

  45. [53]

    Hear me out: Fusional approaches for audio augmented temporal action localization,

    A. Bagchi, J. Mahmood, D. Fernandes, and R. K. Sarvadevabhatla, “Hear me out: Fusional approaches for audio augmented temporal action localization,”arXiv preprint arXiv:2106.14118, 2021

    2021 arXiv

  46. [54]

    wav2vec 2.0: A framework for self-supervised learning of speech representations,

    A. Baevski, Y . Zhou, A. Mohamed, and M. Auli, “wav2vec 2.0: A framework for self-supervised learning of speech representations,” Advances in neural information processing systems, vol. 33, pp. 12 449– 12 460, 2020

    2020

  47. [55]

    Vigo: Audiovisual fake detection and segment localization,

    D. P ´erez-Vieites, J. J. Moreira-P ´erez, ´A. Arag ´on-Kifute, R. Rom ´an- Sarmiento, and R. Castro-Gonz ´alez, “Vigo: Audiovisual fake detection and segment localization,” inProceedings of the 32nd ACM Interna- tional Conference on Multimedia, 2024, pp. 11 360–11 364

    2024

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.