Pith. sign in

REVIEW 3 major objections 5 minor 37 references

Towards Open-world Generalized Deepfake Detection: General Feature Extraction via Unsupervised Domain Adaptation

T0 review · 3 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read This paper claims a training strategy, OWG-DS, that adapts deepfake detectors to new, unlabeled forgery methods using a small labeled source domain, with large AUC gains in cross-method and cross-dataset tests.

desk verdict Strong empirical gains but a load-bearing adversarial mechanism that is unspecified in the main text; fix that and this is a solid conditional accept. read the letter →

arxiv 2505.12339 v1 pith:M7GDB4FF submitted 2025-05-18 cs.CV cs.AI

Midou Guo , Qilin Yin , Wei Lu , Xiangyang Luo This is my paper · ORCID
classification cs.CVcs.AI
keywords deepfakedetectionopen-worldunsuperviseddomainadaptationshiftfeaturealignmentfaceforensicsgeneralizationXception
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

Open-world deepfake detection faces a mismatch: new forgery methods keep appearing, but labels are scarce. The paper defines a task in which a detector has a small labeled source domain and a large unlabeled target domain whose fake methods are disjoint from the source, and proposes OWG-DS, a training strategy that adapts any encoder to the unlabeled domain. It combines domain-centroid alignment, similarity-based class-boundary separation, and an adversarial domain classifier to align features across domains while keeping real and fake separable. If the strategy works as claimed, detectors could be updated against new deepfake methods using raw, unlabeled data, which is exactly the situation social platforms face. The paper reports strong cross-dataset gains, including an AUC rise from 72.33% to 99.51% on Celeb-DF.

What carries the argument

The machinery is a summed training objective over four terms: cross-entropy on labeled source samples, a KL-divergence regularization term $R$, the domain alignment loss $L_{DAL}$, and the similarity loss $L_{SCBS}$. The Domain Distance Optimization module computes momentum-updated global centroids for each domain and a loss that reduces the centroid distance while expanding the average within-domain spread, using a dynamic weight $w_{intra}=1-\text{epoch}/\text{Epoch}$. The Similarity-based Class Boundary Separation module builds positive pairs: same-label pairs in the source domain and second-nearest-neighbor pairs by cosine similarity in the target domain, then maximizes their similarity. The adversarial domain classifier contributes $L_{adv}$ to bridge the two domains. Together these terms are meant to align feature distributions without collapsing the real/fake boundary.

What would settle it

Run the released code and inspect the backward pass for $L_{adv}$: if the encoder receives the ordinary gradient of Eq. (10) with no gradient reversal or equivalent modulation, then the ADC is not adversarial and the reported gains must come from DDO and SCBS, which can be tested by ablating the three terms under a correctly specified adversarial update.

Watch

Extended reading notes

Core claim

The central claim is that a deepfake detector can be made to recognize new, unseen forgery methods by adapting it to a large unlabeled target corpus using only a small labeled source corpus, provided the two corpora's fake sets are disjoint. The paper calls this the open-world deepfake detection generalization task and proposes OWG-DS, a plug-in training strategy for an already pretrained encoder. Feature alignment is driven by a Domain Distance Optimization module that shrinks the Euclidean distance between momentum-smoothed domain centroids while expanding the average intra-domain spread, with loss $L_{DAL}=D_{inter}+\exp(-(D^S_{intra}+D^T_{intra}))w_{intra}$; a Similarity-based Class Boundary Separation module pulls same-class or nearest-neighbor features closer so that real and fake remain separable during alignment; and a binary cross-entropy adversarial domain classifier is meant to render features domain-invariant. On FF++ to Celeb-DF the strategy raises target AUC from 72.33% (Xception baseline) to 99.51%, and on FF++ to DFDC from 65.61% to 89.37%, while largely preserving source-domain accuracy.

Load-bearing premise

The load-bearing premise is that the adversarial domain classifier actually pushes the encoder to discard domain information, but the paper specifies only a domain-label cross-entropy loss whose ordinary gradient would preserve domain information, so the claimed adversarial effect depends on an unstated update rule.

Editorial extensions

If this is right

  • Target-domain detection can be improved against unseen forgery methods without annotating any target frames, which is the data condition social platforms actually face.
  • The strategy is not tied to one architecture: the paper reports consistent target-domain gains with Xception, ResNet-50, and EfficientNet-B0.
  • Adaptation is data-efficient: using only 30% of the target domain data still beats the unadapted baseline by roughly 16 to 20 AUC points in the paper's two tested scenarios.
  • The gains appear in both cross-method (within FF++) and cross-dataset (FF++ to Celeb-DF and DFDC) settings, so the mechanism is not limited to one kind of domain shift.

Reading between the lines

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

  • Because the paper does not specify the gradient update for the adversarial domain classifier, a fair test of the claimed mechanism is to run the released code and check whether the encoder receives a reversed or modified gradient for $L_{adv}$; if not, the reported gains are attributable to the DDO and SCBS terms alone.
  • The pseudo-positive pairing for unlabeled target samples assumes the pretrained encoder's cosine similarity is reliable enough that second-nearest neighbors are usually same-class; a natural stress test is to vary the pretraining epoch or the domain gap and measure whether SCBS starts pairing across classes.
  • The DDO objective expands intra-domain variance while aligning centroids; one extension would be to study whether this aids or hurts fine-grained subclasses of fake data, such as different manipulation intensities, since the paper only reports binary real/fake accuracy.
  • The task definition excludes overlap between source and target fake domains; a practical extension would be to relax that to the partially overlapping case and check whether the same losses still help.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper defines an open-world deepfake detection generalization task in which a model is trained on a small labeled source domain and a large unlabeled target domain, with source and target fake data disjoint. It proposes an enhancement strategy, OWG-DS, built from three components: a Domain Distance Optimization (DDO) module that aligns domain centroids and expands intra-domain spread, a Similarity-based Class Boundary Separation (SCBS) module that pulls together similar samples, and an Adversarial Domain Classifier (ADC) intended to make features domain-invariant. Experiments on FF++, Celeb-DF, and DFDC in cross-manipulation and cross-dataset scenarios report large target-domain AUC gains over Xception, ORCA, DomainForensics, and OSDD, with additional ablations, data-efficiency tests, and backbone-agnostic results.

Significance. If the training objective is implemented as the paper intends, the task formulation is relevant to open-world deepfake detection and the reported empirical gains are substantial. The paper deserves credit for evaluating cross-method and cross-dataset transfer, for testing data efficiency down to 30% of the target domain, and for showing that the strategy can be attached to multiple backbones. However, the central adversarial mechanism is under-specified in a way that directly affects the claimed domain-invariant feature learning, and one loss weight has a sign that contradicts the stated regularization objective. These issues must be resolved before the reported results can be attributed to the proposed method as written.

major comments (3)
  1. [§3.5 and Eq. (10)] The adversarial domain classifier is not actually defined as an adversarial update. Equation (10) is a standard binary cross-entropy loss for source/target domain classification, and the total loss in Eq. (12) adds it with a positive coefficient η3=1. If the encoder is trained to minimize the total loss as written, it is rewarded for making source and target features more separable, not for making them domain-invariant. The phrase 'dynamic gradient modulation' in §3.5 does not specify the update rule; there is no gradient-reversal layer, no alternating optimization, and no negative gradient step for the encoder. Because the ablation in Table 3 treats ADC as a necessary component and Tables 1 and 2 attribute gains to domain alignment, the missing adversarial mechanism is load-bearing for the paper's central claim.
  2. [§3.6 and implementation details in §4.1] Setting η4=-1 in the total loss contradicts the description of the regularization term R. The text states that R is used to keep the model output diverse and approximates maximum-entropy regularization, which requires minimizing the KL divergence toward the prior distribution. With η4=-1, the term is maximized, driving predictions toward a degenerate, peaked class distribution. The authors either need a positive weight for R or a detailed explanation of why maximizing the KL term is intended. This is not a cosmetic hyperparameter choice; it reverses the stated effect of the regularization.
  3. [§3.4 and Eq. (9)] The SCBS loss as written only maximizes cosine similarity of positive pairs; it contains no term that pushes samples from different classes apart. The text in §3.4 says the module pushes apart samples from different classes and sharpens class boundaries, but Eq. (9) has no negative-pair or repulsion term. If separation is achieved indirectly through clustering or through the supervised source loss, that mechanism should be stated explicitly. Additionally, σ(s_ij) is described as a softmax function applied to a scalar similarity; the authors should specify whether this is a sigmoid or a softmax over a defined set of similarities.
minor comments (5)
  1. [Eq. (8)] I checked the alleged sign issue in Eq. (8) and do not find one: because exp(-D) decreases as D increases, minimizing the second term of L_DAL indeed increases the intra-domain distances, which is consistent with the text's goal of expanding intra-domain divergence.
  2. [Table 3] The ablation table is interpretable, but the column headers should indicate that the checkmarks mean 'module included', and the baseline for the Δ column (the full model) should be stated explicitly in the caption or table notes.
  3. [Notation] The symbol y_i is used for real/fake labels in §3.1 but for domain labels in Eq. (10); using d_i for the domain label would avoid ambiguity.
  4. [Hyperparameters] The momentum coefficient μ in Eq. (5), the prior distribution P in Eq. (11), and the exact schedule of w_intra are not specified in the implementation details; these values are needed to reproduce the method.
  5. [Figures and prose] Figure 5 lacks axis labels and a legend, and the caption does not explain the solid/dashed distinction. There are also several typos, e.g., 'a open world', 'KullbackLeiler', and 'effectively detection extensive unlabeled data'.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found; the training strategy is evaluated on independent target-domain labels.

full rationale

The paper proposes OWG-DS, an empirical training strategy combining DDO, SCBS, and ADC, and reports cross-domain deepfake detection performance. The central claims are measured by AUC/ACC on target domains whose labels are never used during adaptation, so the prediction target is not an input to the method. The DDO and SCBS objectives are training losses, not redefinitions of the evaluation metric. SCBS uses a feature-space neighbor screen to form pseudo-positive pairs and then maximizes their similarity; this is a self-training mechanism, but it does not make the reported target-domain accuracy equal to the training objective by construction, because the final evaluation uses ground-truth labels independent of the pseudo-pairing. The regularization term is borrowed from an external citation (ORCA) and is not presented as a derived first-principles result. No self-citation chain or imported uniqueness theorem is load-bearing. The adversarial domain classifier is underspecified (Eq. 10 is a plain cross-entropy loss added with a positive weight, so the paper does not actually demonstrate that minimizing it yields domain-invariant features), but that is a correctness/rigor concern, not a circularity: the stated equations and claims do not reduce to their inputs by definition. There is no fitted parameter renamed as a prediction, and no known result is merely relabeled. The derivation chain is self-contained with respect to circularity.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The method rests on hand-set loss weights, an unspecified momentum coefficient, and an unstated adversarial update rule; these are the main free choices. No new entities are introduced.

free parameters (4)
  • Loss weights eta1, eta2, eta3, eta4 = 0.1, 1, 1, -1
    Hand-set hyperparameters in the final loss (Eq. 12); no tuning procedure or sensitivity analysis is reported.
  • Momentum coefficient mu = not specified
    Used to update global centroids in Eq. (5); the paper does not give its value or sensitivity.
  • Prior distribution P = not specified
    Appears in regularization term R (Eq. 11); the text says max entropy is used, so P is likely uniform but not stated.
  • Dynamic weight w_intra schedule = 1 - epoch/Epoch
    Ad hoc schedule controlling intra-domain divergence; no ablation justifies the specific form.
assumptions (3)
  • ad hoc to paper Reducing inter-domain centroid distance and increasing intra-domain distances improves cross-domain generalization
    The core mechanism of DDO (Section 3.3) is asserted without theoretical or empirical justification that expanding intra-domain spread helps classification.
  • domain assumption Pre-trained feature extractor provides reliable similarity for pseudo-positive pair selection
    SCBS (Section 3.4) relies on nearest-neighbor pairs being mostly same-class; the paper argues this without evidence.
  • standard math Standard Euclidean and cosine geometry in feature space
    Not explicitly stated but assumed in all distance and similarity calculations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Towards Open-world Generalized Deepfake Detection: General Feature Extraction via Unsupervised Domain Adaptation." pith.science (2026). https://pith.science/paper/M7GDB4FF

@misc{pith2026250512339,
  author       = {Pith},
  title        = {Pith review of: Towards Open-world Generalized Deepfake Detection: General Feature Extraction via Unsupervised Domain Adaptation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/M7GDB4FF}},
  note         = {Machine review of arXiv:2505.12339}
}
read the original abstract

With the development of generative artificial intelligence, new forgery methods are rapidly emerging. Social platforms are flooded with vast amounts of unlabeled synthetic data and authentic data, making it increasingly challenging to distinguish real from fake. Due to the lack of labels, existing supervised detection methods struggle to effectively address the detection of unknown deepfake methods. Moreover, in open world scenarios, the amount of unlabeled data greatly exceeds that of labeled data. Therefore, we define a new deepfake detection generalization task which focuses on how to achieve efficient detection of large amounts of unlabeled data based on limited labeled data to simulate a open world scenario. To solve the above mentioned task, we propose a novel Open-World Deepfake Detection Generalization Enhancement Training Strategy (OWG-DS) to improve the generalization ability of existing methods. Our approach aims to transfer deepfake detection knowledge from a small amount of labeled source domain data to large-scale unlabeled target domain data. Specifically, we introduce the Domain Distance Optimization (DDO) module to align different domain features by optimizing both inter-domain and intra-domain distances. Additionally, the Similarity-based Class Boundary Separation (SCBS) module is used to enhance the aggregation of similar samples to ensure clearer class boundaries, while an adversarial training mechanism is adopted to learn the domain-invariant features. Extensive experiments show that the proposed deepfake detection generalization enhancement training strategy excels in cross-method and cross-dataset scenarios, improving the model's generalization.

Figures

Figures reproduced from arXiv: 2505.12339 by the authors.

Figure 1
Figure 1. Illustration of different deepfake detection tasks: (a) Traditional detection relies on labeled data to classify data as [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The framework of the Open-World Generalization Enhancement Training Strategy (OWG-DS) includes the Domain [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. The Domain Distance Optimization (DDO) diagram shows the dynamic process of cross-domain feature alignment [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: T-SNE visualization on FF++ (FF FS NT → DF) (HQ). domain is constrained across all baseline methods, likely due to the significant distribution gap between the source and target domains, which increases the difficulty of domain adaptation. However, even under this chal…
Figure 5
Figure 5. Figure 5: Data efficiency on (DF FS NT-FF) and (FF FS NT-DF) [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

37 extracted references · 25 canonical work pages

  1. [1]

    In ictu oculi: Exposing ai generated fake face videos by detecting eye blinking

    Yuezun Li, Ming-Ching Chang, and Siwei Lyu. In ictu oculi: Exposing ai generated fake face videos by detecting eye blinking. arXiv preprint arXiv:1806.02877, 2018

    arXiv 2018

  2. [2]

    Mesonet: a compact facial video forgery detection network

    Darius Afchar, Vincent Nozick, Junichi Yamagishi, and Isao Echizen. Mesonet: a compact facial video forgery detection network. In 2018 IEEE international workshop on information forensics and security (WIFS) , pages 1–7. IEEE, 2018

    work page 2018

  3. [3]

    Frequency-aware discriminative feature learning supervised by single-center loss for face forgery detection

    Jiaming Li, Hongtao Xie, Jiahong Li, Zhongyuan Wang, and Yongdong Zhang. Frequency-aware discriminative feature learning supervised by single-center loss for face forgery detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 6458–6467, 2021

    work page 2021

  4. [4]

    Improving the efficiency and robustness of deepfakes detection through precise geometric features

    Zekun Sun, Yujie Han, Zeyu Hua, Na Ruan, and Weijia Jia. Improving the efficiency and robustness of deepfakes detection through precise geometric features. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 3609–3618, 2021

    work page 2021

  5. [5]

    Forgerynet: A versatile benchmark for comprehensive forgery analysis

    Yinan He, Bei Gan, Siyu Chen, Yichun Zhou, Guojun Yin, Luchuan Song, Lu Sheng, Jing Shao, and Ziwei Liu. Forgerynet: A versatile benchmark for comprehensive forgery analysis. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 4360–4369, 2021

    work page 2021

  6. [6]

    Altfreezing for more general video face forgery detection

    Zhendong Wang, Jianmin Bao, Wengang Zhou, Weilun Wang, and Houqiang Li. Altfreezing for more general video face forgery detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 4129–4138, 2023

    work page 2023

  7. [7]

    Tall: Thumbnail layout for deepfake video detection

    Yuting Xu, Jian Liang, Gengyun Jia, Ziming Yang, Yanhao Zhang, and Ran He. Tall: Thumbnail layout for deepfake video detection. In Proceedings of the IEEE/CVF international conference on computer vision , pages 22658–22668, 2023

    work page 2023

  8. [8]

    Faceforensics++: Learning to detect manipulated facial images

    Andreas Rossler, Davide Cozzolino, Luisa Verdoliva, Christian Riess, Justus Thies, and Matthias Nießner. Faceforensics++: Learning to detect manipulated facial images. In Proceedings of the IEEE/CVF international conference on computer vision, pages 1–11, 2019

    work page 2019

Show all 37 references
  1. [9]

    Face x-ray for more general face forgery detection

    Lingzhi Li, Jianmin Bao, Ting Zhang, Hao Yang, Dong Chen, Fang Wen, and Baining Guo. Face x-ray for more general face forgery detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 5001–5010, 2020

    2020

  2. [10]

    Denoising diffusion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffusion probabilistic models. Advances in neural information processing systems , 33:6840–6851, 2020

    2020

  3. [11]

    Exposing deepfake videos by detecting face warping artifacts

    Yuezun Li and Siwei Lyu. Exposing deepfake videos by detecting face warping artifacts. arXiv preprint arXiv:1811.00656, 2018

    2018 arXiv

  4. [12]

    Learning self-consistency for deepfake detection

    Tianchen Zhao, Xiang Xu, Mingze Xu, Hui Ding, Yuanjun Xiong, and Wei Xia. Learning self-consistency for deepfake detection. In Proceedings of the IEEE/CVF international conference on computer vision , pages 15023–15033, 2021

    2021

  5. [13]

    Detecting deepfakes with self-blended images

    Kaede Shiohara and Toshihiko Yamasaki. Detecting deepfakes with self-blended images. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 18720–18729, 2022

    2022

  6. [14]

    Self- supervised learning of adversarial example: Towards good generalizations for deepfake detection

    Liang Chen, Yong Zhang, Yibing Song, Lingqiao Liu, and Jue Wang. Self- supervised learning of adversarial example: Towards good generalizations for deepfake detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 18710–18719, 2022

    2022

  7. [15]

    Lisiam: Localization invariance siamese network for deepfake detection

    Jian Wang, Yunlian Sun, and Jinhui Tang. Lisiam: Localization invariance siamese network for deepfake detection. IEEE Transactions on Information Forensics and Security, 17:2425–2436, 2022

    2022

  8. [16]

    Dire for diffusion-generated image detection

    Zhendong Wang, Jianmin Bao, Wengang Zhou, Weilun Wang, Hezhen Hu, Hong Chen, and Houqiang Li. Dire for diffusion-generated image detection. In Pro- ceedings of the IEEE/CVF International Conference on Computer Vision , pages 22445–22455, 2023

    2023

  9. [17]

    Forensictransfer: Weakly-supervised domain adaptation for forgery detection

    Davide Cozzolino, Justus Thies, Andreas Rössler, Christian Riess, Matthias Nießner, and Luisa Verdoliva. Forensictransfer: Weakly-supervised domain adaptation for forgery detection. arXiv preprint arXiv:1812.02510, 2018

    2018 arXiv

  10. [18]

    Generalized zero and few-shot transfer for facial forgery detection

    Shivangi Aneja and Matthias Nießner. Generalized zero and few-shot transfer for facial forgery detection. arXiv preprint arXiv:2006.11863, 2020

    2006 arXiv

  11. [19]

    Few-shot forgery detection via guided adversarial interpolation

    Haonan Qiu, Siyu Chen, Bei Gan, Kun Wang, Huafeng Shi, Jing Shao, and Ziwei Liu. Few-shot forgery detection via guided adversarial interpolation. Pattern Recognition, 144:109863, 2023

    2023

  12. [20]

    Uni- fied deep supervised domain adaptation and generalization

    Saeid Motiian, Marco Piccirilli, Donald A Adjeroh, and Gianfranco Doretto. Uni- fied deep supervised domain adaptation and generalization. In Proceedings of the IEEE international conference on computer vision , pages 5715–5725, 2017

    2017

  13. [21]

    Identity-driven multimedia forgery detection via reference assistance

    Junhao Xu, Jingjing Chen, Xue Song, Feng Han, Haijun Shan, and Yu-Gang Jiang. Identity-driven multimedia forgery detection via reference assistance. In Proceedings of the 32nd ACM International Conference on Multimedia , pages 3887–3896, 2024

    2024

  14. [22]

    Temporal ensembling for semi-supervised learning

    Samuli Laine and Timo Aila. Temporal ensembling for semi-supervised learning. arXiv preprint arXiv:1610.02242, 2016

    2016 arXiv

  15. [23]

    Open-world semi-supervised learning

    Kaidi Cao, Maria Brbic, and Jure Leskovec. Open-world semi-supervised learning. arXiv preprint arXiv:2102.03526, 2021

    2021 arXiv

  16. [24]

    Virtual adver- sarial training: a regularization method for supervised and semi-supervised learn- ing

    Takeru Miyato, Shin-ichi Maeda, Masanori Koyama, and Shin Ishii. Virtual adver- sarial training: a regularization method for supervised and semi-supervised learn- ing. IEEE transactions on pattern analysis and machine intelligence , 41(8):1979– 1993, 2018

    1979

  17. [25]

    Pseudo-label: The simple and efficient semi-supervised learning method for deep neural networks

    Dong-Hyun Lee et al. Pseudo-label: The simple and efficient semi-supervised learning method for deep neural networks. In Workshop on challenges in repre- sentation learning, ICML, volume 3, page 896. Atlanta, 2013

    2013

  18. [26]

    Unsupervised domain adaptation for semantic segmentation via class-balanced self-training

    Yang Zou, Zhiding Yu, BVK Kumar, and Jinsong Wang. Unsupervised domain adaptation for semantic segmentation via class-balanced self-training. In Pro- ceedings of the European conference on computer vision (ECCV) , pages 289–305, 2018

    2018

  19. [27]

    Adversarial discrim- inative domain adaptation

    Eric Tzeng, Judy Hoffman, Kate Saenko, and Trevor Darrell. Adversarial discrim- inative domain adaptation. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 7167–7176, 2017

    2017

  20. [28]

    Unsupervised domain adaptation by backpropagation

    Yaroslav Ganin and Victor Lempitsky. Unsupervised domain adaptation by backpropagation. In International conference on machine learning , pages 1180–

  21. [29]

    Separate to adapt: Open set domain adaptation via progressive separation

    Hong Liu, Zhangjie Cao, Mingsheng Long, Jianmin Wang, and Qiang Yang. Separate to adapt: Open set domain adaptation via progressive separation. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 2927–2936, 2019

    2019

  22. [30]

    Domainforensics: Exposing face forgery across domains via bi-directional adaptation

    Qingxuan Lv, Yuezun Li, Junyu Dong, Sheng Chen, Hui Yu, Huiyu Zhou, and Shu Zhang. Domainforensics: Exposing face forgery across domains via bi-directional adaptation. IEEE Transactions on Information Forensics and Security , 2024

    2024

  23. [31]

    Fine-grained open-set deepfake detection via unsupervised domain adaptation

    Xinye Zhou, Hu Han, Shiguang Shan, and Xilin Chen. Fine-grained open-set deepfake detection via unsupervised domain adaptation. IEEE Transactions on Information Forensics and Security , 2024

    2024

  24. [32]

    Celeb-df: A large- scale challenging dataset for deepfake forensics

    Yuezun Li, Xin Yang, Pu Sun, Honggang Qi, and Siwei Lyu. Celeb-df: A large- scale challenging dataset for deepfake forensics. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 3207–3216, 2020

    2020

  25. [33]

    The deepfake detection challenge (dfdc) dataset

    Brian Dolhansky, Joanna Bitton, Ben Pflaum, Jikuo Lu, Russ Howes, Menglin Wang, and Cristian Canton Ferrer. The deepfake detection challenge (dfdc) dataset. arXiv preprint arXiv:2006.07397, 2020

    2006 arXiv

  26. [34]

    Xception: Deep learning with depthwise separable convolutions

    François Chollet. Xception: Deep learning with depthwise separable convolutions. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 1251–1258, 2017

    2017

  27. [35]

    Frequency-aware deepfake detection: Improving generalizability through frequency space domain learning

    Chuangchuang Tan, Yao Zhao, Shikui Wei, Guanghua Gu, Ping Liu, and Yunchao Wei. Frequency-aware deepfake detection: Improving generalizability through frequency space domain learning. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 5052–5060, 2024

    2024

  28. [36]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016

    2016

  29. [37]

    Efficientnet: Rethinking model scaling for convolu- tional neural networks

    Mingxing Tan and Quoc Le. Efficientnet: Rethinking model scaling for convolu- tional neural networks. In International conference on machine learning , pages 6105–6114. PMLR, 2019

    2019

Pith tools

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