DX's Digital Garden

论文

10 min read

[2] Xie T, Yu L, Luo C, et al. Survey of deep face manipulation and fake detection[J]. Journal of Tsinghua University (Science and Technology), 2023, 63(9): 1350-1365.
DOI: 10.16511/j.cnki.qhdxxb.2023.21.002
链接: SciOpen

二、早期基于生理特征与几何一致性的检测方法
[3] Li Y, Chang M C, Lyu S. In Ictu Oculi: Exposing AI generated fake face videos by detecting eye blinking[EB/OL]. arXiv, 2018.
DOI: 10.48550/arXiv.1806.02877
链接: arXiv

[4] Yang X, Li Y, Lyu S. Exposing deep fakes using inconsistent head poses[C]//ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing. IEEE, 2019: 8261-8265.
DOI: 10.1109/ICASSP.2019.8683164
链接: SUNY Research Connect

三、基于空间域与卷积网络的检测方法
[5] Afchar D, Nozick V, Yamagishi J, et al. MesoNet: A compact facial video forgery detection network[C]//2018 IEEE International Workshop on Information Forensics and Security (WIFS). IEEE, 2018: 1-7.
DOI: 10.1109/WIFS.2018.8630761
链接: arXiv

[6] Rossler A, Cozzolino D, Verdoliva L, et al. FaceForensics++: Learning to detect manipulated facial images[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). 2019: 1-11.
DOI: 10.1109/ICCV.2019.00009
链接: CVF Open Access

[7] Dang H, Liu F, Stehouwer J, et al. On the detection of digital face manipulation[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2020: 5781-5790.
DOI: 10.1109/CVPR42600.2020.00582
链接: CVF Open Access

[8] Li L, Bao J, Zhang T, et al. Face X-ray for more general face forgery detection[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2020: 5001-5010.
DOI: 10.1109/CVPR42600.2020.00506
链接: CVF Open Access

[9] Chollet F. Xception: Deep learning with depthwise separable convolutions[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 2017: 1251-1258.
DOI: 10.1109/CVPR.2017.195
链接: DOI
说明：如果你正文里保留“代表性方法如 Xception”这句话，建议把这篇也一起引上。

四、基于频域特征的检测方法
[10] Qian Y, Yin G, Sheng L, et al. Thinking in frequency: Face forgery detection by mining frequency-aware clues[C]//Computer Vision – ECCV 2020. Springer, 2020: 86-103.
DOI: 10.1007/978-3-030-58610-2_6
链接: Springer/DOI

[11] Liu H, Li X, Zhou W, et al. Spatial-phase shallow learning: Rethinking face forgery detection in frequency domain[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2021: 772-781.
DOI: 10.1109/CVPR46437.2021.00083
链接: DOI

五、基于时序一致性与视频动态特征的检测方法
[12] Haliassos A, Vougioukas K, Petridis S, et al. Lips don’t lie: A generalisable and robust approach to face forgery detection[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2021: 5039-5049.
DOI: 10.1109/CVPR46437.2021.00500
链接: CVF Open Access

[13] Zheng Y, Bao J, Chen D, et al. Exploring temporal coherence for more general video face forgery detection[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). 2021: 15044-15054.
DOI: 10.1109/ICCV48922.2021.01477
链接: CVF Open Access

[14] Haliassos A, Mira R, Petridis S, et al. Leveraging real talking faces via self-supervision for robust forgery detection[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2022: 14950-14962.
DOI: 10.1109/CVPR52688.2022.01453
链接: CVF Open Access

六、基于重建与多线索建模的检测方法
[15] Cao J, Ma C, Yao T, et al. End-to-end reconstruction-classification learning for face forgery detection[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2022: 4113-4122.
DOI: 10.1109/CVPR52688.2022.00408
链接: CVF Open Access

如果你想严格对应我刚才那版 1.2.2 的文字，最少保留这 10 篇就够了

  • [3][4] 对应“手工特征、生理特征、几何一致性”
  • [5][6][7][8] 对应“空间域与卷积网络方法”
  • [10][11] 对应“频域方法”
  • [12][13][14] 对应“时序一致性与视频方法”
  • [15] 对应“重建与多线索融合方法” 1.2.3 面向泛化的深度伪造检测研究现状
    本段可用参考文献
    [1] Shiohara K, Yamasaki T. Detecting Deepfakes With Self-Blended Images[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2022: 18720-18729.
    链接: https://openaccess.thecvf.com/content/CVPR2022/html/Shiohara_Detecting_Deepfakes_With_Self-Blended_Images_CVPR_2022_paper.html

[2] Yan Z, Zhang Y, Fan Y, et al. UCF: Uncovering Common Features for Generalizable Deepfake Detection[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision. 2023: 22412-22423.
链接: https://openaccess.thecvf.com/content/ICCV2023/html/Yan_UCF_Uncovering_Common_Features_for_Generalizable_Deepfake_Detection_ICCV_2023_paper.html

[3] Yan Z, Luo Y, Lyu S, et al. Transcending Forgery Specificity with Latent Space Augmentation for Generalizable Deepfake Detection[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2024: 8984-8994.
链接: https://openaccess.thecvf.com/content/CVPR2024/html/Yan_Transcending_Forgery_Specificity_with_Latent_Space_Augmentation_for_Generalizable_Deepfake_CVPR_2024_paper.html

[4] Fu X, Yan Z, Yao T, et al. Exploring Unbiased Deepfake Detection via Token-Level Shuffling and Mixing[C]//Proceedings of the AAAI Conference on Artificial Intelligence. 2025, 39(3): 3040-3048.
DOI: https://doi.org/10.1609/aaai.v39i3.32312

[5] Radford A, Kim J W, Hallacy C, et al. Learning Transferable Visual Models From Natural Language Supervision[C]//Proceedings of the 38th International Conference on Machine Learning. PMLR, 2021: 8748-8763.
链接: https://proceedings.mlr.press/v139/radford21a.html

[6] Lin K, Lin Y, Li W, et al. Standing on the Shoulders of Giants: Reprogramming Visual-Language Model for General Deepfake Detection[C]//Proceedings of the AAAI Conference on Artificial Intelligence. 2025, 39(5): 5262-5270.
DOI: https://doi.org/10.1609/aaai.v39i5.32559

[7] Cui X, Li Y, Luo A, et al. Forensics Adapter: Adapting CLIP for Generalizable Face Forgery Detection[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2025: 19207-19217.
链接: https://openaccess.thecvf.com/content/CVPR2025/html/Cui_Forensics_Adapter_Adapting_CLIP_for_Generalizable_Face_Forgery_Detection_CVPR_2025_paper.html

[8] Guo X, Song X, Zhang Y, et al. Rethinking Vision-Language Model in Face Forensics: Multi-Modal Interpretable Forged Face Detector[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2025: 105-116.
链接: https://openaccess.thecvf.com/content/CVPR2025/html/Guo_Rethinking_Vision-Language_Model_in_Face_Forensics_Multi-Modal_Interpretable_Forged_Face_CVPR_2025_paper.html

[9] Zhang Y, Wang T, Yu Z, et al. MFCLIP: Multi-Modal Fine-Grained CLIP for Generalizable Diffusion Face Forgery Detection[J]. IEEE Transactions on Information Forensics and Security, 2025, 20: 5888-5903.
DOI: https://doi.org/10.1109/TIFS.2025.3576577
链接: https://research.nottingham.edu.cn/en/publications/mfclip-multi-modal-fine-grained-clip-for-generalizable-diffusion- 1.2.4 基于视觉基础模型/CLIP 的深度伪造检测研究现状
1.3 现有研究存在的问题
1.4 本文主要研究内容
1.5 本文组织架构

第二章 相关理论与关键技术
2.1 深度伪造检测任务定义与挑战
2.1.1 伪造类型与检测任务
2.1.2 跨数据集检测与泛化问题
2.2 深度学习基础
2.2.1 卷积神经网络
2.2.2 Transformer
2.2.3 注意力机制
2.3 CLIP模型相关理论
2.3.1 CLIP预训练机制
2.3.2 图像编码器与视觉表征
2.3.3 CLIP在取证任务中的适配方式
2.4 适配器学习相关技术
2.4.1 参数高效微调思想
2.4.2 并行适配器结构
2.4.3 面向取证任务的适配器设计
2.5 本章小结

第三章 基于聚焦-遗忘适配框架的深度伪造检测方法
这一章就是把你现在的 F-Adapter 完整中文化、论文化。

3.1 问题分析与方法动机
3.2 整体框架设计
3.3 伪影聚焦模块
3.3.1 区域划分与特征投影
3.3.2 区域级稀疏路由机制
3.3.3 基于内容感知的细粒度注意力
3.4 捷径遗忘模块
3.4.1 特征响应重要性估计
3.4.2 课程式空间掩码策略
3.4.3 软到硬的抑制机制
3.5 模型优化目标
3.5.1 真实性分类损失
3.5.2 边界重建损失
3.5.3 Patch级对比损失
3.5.4 Sample级对比损失
3.6 实验设置
3.6.1 数据集与预处理
3.6.2 评价指标
3.6.3 对比方法
3.6.4 实现细节
3.7 实验结果与分析
3.7.1 跨数据集帧级检测结果
3.7.2 跨数据集视频级检测结果
3.7.3 多伪造类型检测结果
3.8 消融实验
3.8.1 聚焦模块有效性分析
3.8.2 遗忘模块有效性分析
3.8.3 稀疏路由参数分析
3.8.4 遗忘策略分析
3.9 扩展分析
3.9.1 鲁棒性分析
3.9.2 可视化分析
3.9.3 计算复杂度分析
3.10 本章小结

第四章 基于CLIP适配器的深度伪造检测方法
这一章先保留，不把技术路线写死。

4.1 问题分析与研究动机
4.2 整体框架设计
4.3 核心适配器模块设计
4.4 分类判别策略设计
4.5 优化目标
4.6 实验设置
4.7 实验结果与分析
4.8 消融实验
4.9 本章小结

Graph View