不确定性感知的多视角结构学习深度伪造检测
原标题：Uncertainty-Aware Deepfake Detection via Multi-View Structural Learning
AI 摘要
该研究提出了一种基于多视角结构学习的不确定性感知深度伪造检测框架，通过视觉、语义和结构三个流整合互补证据，并引入分支间分歧校准（IBDC）机制将预测不确定性与证据冲突关联。在FaceForensics++训练集上的跨数据集实验表明，该方法在多个分布外基准上实现了最先进的泛化性能，同时改善了校准和选择性预测能力，为分布偏移下的可信深度伪造检测提供了稳健基础。
正文节选
Computer Science > Computer Vision and Pattern Recognition Title:Uncertainty-Aware Deepfake Detection via Multi-View Structural Learning View PDF HTML (experimental) Abstract:Security-critical biometric and forensic applications require accurate predictions and reliable confidence estimates, particularly under distribution shift. This challenge is especially acute for deepfake detection, where foundation-model-based detectors often exhibit overconfident predictions on out-of-distribu