构建更贴近真实伪造的训练模型，提升深度伪造检测鲁棒性。
Practical Manipulation Model for Robust Deepfake Detection
- 用泊松融合、多样掩码等生成更多逼真伪造图像
- 在标准数据集上检测性能提升3.51%至6.21% AUC
- 适合关注实际场景下检测可靠性的研究人员
当前深度伪造检测模型在跨数据集任务中表现优异，但在非理想条件下仍不稳定，易被绕过。受图像超分辨率领域向真实退化建模的启发，我们提出实用伪造建模（PMM），涵盖更广泛的伪造可能性。通过泊松融合、多样化掩码、生成器伪影和干扰物扩展伪伪造空间，并在训练图像中加入强退化以提升检测器的泛化性和鲁棒性。实验表明，该方法不仅显著增强对常见图像退化的鲁棒性，还在标准基准上实现性能提升：在DFDC和DFDCP数据集上分别较s-o-t-a的LAA主干网络提升3.51%和6.21% AUC。同时揭示了以往检测器的脆弱性，并在此方面取得改进。代码见https://github.com/BenediktHopf/PMM。
原文摘要 · Abstract (English)
Modern deepfake detection models have achieved strong performance even on the challenging cross-dataset task. However, detection performance under non-ideal conditions remains very unstable, limiting success on some benchmark datasets and making it easy to circumvent detection. Inspired by the move to a more real-world degradation model in the area of image super-resolution, we have developed a Practical Manipulation Model (PMM) that covers a larger set of possible forgeries. We extend the space of pseudo-fakes by using Poisson blending, more diverse masks, generator artifacts, and distractors. Additionally, we improve the detectors' generality and robustness by adding strong degradations to the training images. We demonstrate that these changes not only significantly enhance the model's robustness to common image degradations but also improve performance on standard benchmark datasets. Specifically, we show clear increases of $3.51\%$ and $6.21\%$ AUC on the DFDC and DFDCP datasets, respectively, over the s-o-t-a LAA backbone. Furthermore, we highlight the lack of robustness in previous detectors and our improvements in this regard. Code can be found at https://github.com/BenediktHopf/PMM
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。