CoChat
Sign Up
Cited by 59Digital Media Forensic DetectionFace recognition and analysisGenerative Adversarial Networks and Image Synthesis

Exploring Disentangled Content Information for Face Forgery Detection

Open access
Liang, Jiahao, Shi, Huafeng, Deng, Weihong·2022·Lecture Notes in Computer Science
DOI 10.1007/978-3-031-19781-9_8
Open this paper in CoChat — free accountOpen PDF

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1007/978-3-031-19781-9_8

Cites

43 of 43 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Disentanglement methods generally underperform compared to other categories, with the best method achieving only 80.5% ACC in cross-domain evaluation.

    “Disentanglement methods [175, 177, 178] generally underperform compared to other categories, with the best method [178] achieving only 80 . 5% ACC in CD evaluation.”

    From Deepfake detection across image, video, and audio: a comprehensive survey with empirical evaluation of generalization and robustness (2026) · cites this paper as [175]

    Section 4.2 (pp. 9–10) reports the paper’s best cross-dataset result as 82.38% AUC, not 80.5% ACC. It also says the proposed disentanglement framework outperforms the compared methods on Celeb-DF, contrary to the claim that this method underperforms other categories. (LLM, conf=0.85)

    Checked by a reasoning model — may be wrong.

    Experimental — accuracy not yet measured

  • Deepfake detectors may inadvertently learn representations containing both forgery-related and content-related information.

    “Disentanglement learning addresses the issue that deepfake detectors may inadvertently learn representations containing both forgery-related and content-related information.”

    From Evolving from Single-modal to Multi-modal Facial Deepfake Detection: Progress and Challenges (2024) · cites this paper as [176]

  • Mixed representations containing both forgery-related and content-related information often lead detectors to overfit by relying on irrelevant cues like face identity or background characteristics.

    “Such mixed representations often lead detectors to overfit by relying on irrelevant cues like face identity or background characteristics [176], [177].”

    From Evolving from Single-modal to Multi-modal Facial Deepfake Detection: Progress and Challenges (2024) · cites this paper as [176]

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

Flagged first, then by citations

Research works printed but not linked: 12

  1. [4]4. Asnani, V., Yin, X., Hassner, T., Liu, X.: Reverse engineering of generative models: Inferring model hyperparameters from generated images. arXiv preprint arXiv:2106.07873 (2021) · arXiv
  2. [11]11. Ciftci, U.A., Demir, I., Yin, L.: Fakecatcher: Detection of synthetic portrait videos using biological signals. IEEE Transactions on Pattern Analysis and Machine Intelligence (2020)
  3. [15]15. Frank, J., Eisenhofer, T., Sch¨ onherr, L., Fischer, A., Kolossa, D., Holz, T.: Leveraging frequency analysis for deep fake image recognition. In: International Conference on Machine Learning. pp. 3247-3258. PMLR (2020)
  4. [19]19. Kingma, D.P., Ba, J.: Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980 (2014) · arXiv
  5. [26]26. Liu, Y., Stehouwer, J., Liu, X.: On disentangling spoof trace for generic face antispoofing. In: European Conference on Computer Vision. pp. 406-422. Springer (2020)
  6. [28]28. Luo, Y., Zhang, Y., Yan, J., Liu, W.: Generalizing face forgery detection with highfrequency features. In: ProceeRings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 16317-16326 (2021)
  7. [29]29. Van der Maaten, L., Hinton, G.: Visualizing data using t-sne. Journal of machine learning research 9 (11) (2008)
  8. [31]31. Masi, I., Killekar, A., Mascarenhas, R.M., Gurudatt, S.P., AbdAlmageed, W.: Twobranch recurrent network for isolating deepfakes in videos. In: European Conference on Computer Vision. pp. 667-684. Springer (2020)
  9. [34]34. Oord, A.v.d., Li, Y., Vinyals, O.: Representation learning with contrastive predictive coding. arXiv preprint arXiv:1807.03748 (2018) · arXiv
  10. [43]43. Tran, L., Yin, X., Liu, X.: Disentangled representation learning gan for poseinvariant face recognition. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 1415-1424 (2017)
  11. [49]49. Zhang, H., Wu, C., Zhang, Z., Zhu, Y., Lin, H., Zhang, Z., Sun, Y., He, T., Mueller, J., Manmatha, R., et al.: Resnest: Split-attention networks. arXiv preprint arXiv:2004.08955 (2020) · arXiv
  12. [50]50. Zhang, H., Cisse, M., Dauphin, Y.N., Lopez-Paz, D.: mixup: Beyond empirical risk minimization. In: International Conference on Learning Representations (2018)

Cited by

Sorted by citations

Related

Authors
Liang, J., Shi, H., & Deng, W.
Affiliations
  • Beijing University of Posts and Telecommunications (CN)
  • Sensetime (China) (CN)
Corresponding author
Deng, Weihong
ORCID
Venue
Lecture Notes in Computer Science
Publisher
Springer Nature Switzerland
Published
2022
Pages
128-145
ISSN
  • 0302-9743 (ISSN-L)
  • 1611-3349
DOI
10.1007/978-3-031-19781-9_8
OpenAlex
W4312516166
Cited by
58
References
43
Citations by year
2026: 8 · 2025: 30 · 2024: 17 · 2023: 3
FWCI
4.42
Type
conference paper
Language
en
Open access
Licence
Locations
  • Lecture notes in computer science (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • arXiv (Cornell University)
Topics
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Digital Media Forensic Detection
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Face recognition and analysis
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Generative Adversarial Networks and Image Synthesis
Keywords
face forgery detection, disentangled representation learning, content bias, overfitting, class imbalance

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review