发布日期: 2025-03-22

更新日期: 2025-03-22

文章字数: 1k

阅读时长: 4 分

阅读次数:

⚠️ 以下所有内容总结都来自于大语言模型的能力，如有错误，仅供参考，谨慎使用
🔴 请注意：千万不要用于严肃的学术场景，只能用于论文阅读前的初筛！
💗 如果您觉得我们的项目对您有帮助 ChatPaperFree ，还请您给我们一些鼓励！⭐️ HuggingFace免费体验

2025-03-22 更新

TruthLens: Explainable DeepFake Detection for Face Manipulated and Fully Synthetic Data

Authors:Rohit Kundu, Athula Balachandran, Amit K. Roy-Chowdhury

Detecting DeepFakes has become a crucial research area as the widespread use of AI image generators enables the effortless creation of face-manipulated and fully synthetic content, yet existing methods are often limited to binary classification (real vs. fake) and lack interpretability. To address these challenges, we propose TruthLens, a novel and highly generalizable framework for DeepFake detection that not only determines whether an image is real or fake but also provides detailed textual reasoning for its predictions. Unlike traditional methods, TruthLens effectively handles both face-manipulated DeepFakes and fully AI-generated content while addressing fine-grained queries such as “Does the eyes/nose/mouth look real or fake?” The architecture of TruthLens combines the global contextual understanding of multimodal large language models like PaliGemma2 with the localized feature extraction capabilities of vision-only models like DINOv2. This hybrid design leverages the complementary strengths of both models, enabling robust detection of subtle manipulations while maintaining interpretability. Extensive experiments on diverse datasets demonstrate that TruthLens outperforms state-of-the-art methods in detection accuracy (by 2-14%) and explainability, in both in-domain and cross-data settings, generalizing effectively across traditional and emerging manipulation techniques.

检测DeepFakes已经成为一个关键的研究领域，随着人工智能图像生成器的广泛应用，可以轻松创建面部操作和完全合成的内容。然而，现有的方法通常仅限于二元分类（真实与虚假），并且缺乏可解释性。为了应对这些挑战，我们提出了TruthLens，这是一个用于DeepFake检测的新型高度通用框架。它不仅确定图像是真实的还是虚假的，而且还为预测提供详细的文本理由。与传统的检测方法不同，TruthLens可以有效地处理面部操作的DeepFakes和完全AI生成的内容，同时处理精细查询，例如“眼睛/鼻子/嘴巴看起来是真实还是假的？”TruthLens的架构结合了多模态大型语言模型（如PaliGemma2）的全局上下文理解与仅视觉模型（如DINOv2）的局部特征提取能力。这种混合设计利用了两种模型的互补优势，能够在保持可解释性的同时，实现微妙的操纵的稳健检测。在多种数据集上的广泛实验表明，TruthLens在检测准确率（高出2-14%）和可解释性方面超过了最先进的方法，并且在域内和跨数据设置中都能有效推广，适用于传统和新兴的操作技术。

论文及项目相关链接

PDF

Summary
真理透镜（TruthLens）框架用于深度伪造检测，不仅能判断图像真伪，还能提供详细的文本推理依据。该框架结合了多模态大型语言模型（如PaliGemma2）的全局上下文理解和仅视觉模型（如DINOv2）的局部特征提取能力，实现精细操作的稳健检测并保持可解释性。在多种数据集上的实验表明，TruthLens在检测准确率上优于最新技术（提高2-14%），且在域内和跨数据设置中的解释性也更强，能够有效地泛化传统和新兴的操纵技术。

Key Takeaways