人类与AI协作可显著提升低质量视频的伪造检测效果
Human-AI Ensembles Improve Deepfake Detection in Low-to-Medium Quality Videos
- 构建人机协同检测系统，利用互补错误模式提升整体性能
- 在手机拍摄的低质视频上，AI准确率降至0.537（接近随机），人类仍保持0.784
- 适合关注真实场景下深度伪造检测的从业者与研究者
深度伪造检测常被视为机器学习问题，但人类与AI在真实条件下的表现对比仍不清晰。我们评估了200名参与者和95个前沿AI检测器在两个数据集上的表现：标准基准DF40和新提出的日常生活视频数据集CharadesDF。CharadesDF由手机录制，画质为低至中等，相比专业拍摄的DF40更具现实性。结果显示，人类在两个数据集上均优于AI，尤其在CharadesDF上差距扩大——AI准确率跌至0.537（接近随机水平），而人类仍维持0.784的稳健表现。人类与AI的错误具有互补性：人类易漏检高质量伪造，而AI常将真实视频误判为伪造。通过融合人机判断，可有效降低高置信度错误。研究表明，在非专业制作视频场景中，深度伪造检测应依赖人机协同而非单一算法。
原文摘要 · Abstract (English)
Deepfake detection is widely framed as a machine learning problem, yet how humans and AI detectors compare under realistic conditions remains poorly understood. We evaluate 200 human participants and 95 state-of-the-art AI detectors across two datasets: DF40, a standard benchmark, and CharadesDF, a novel dataset of videos of everyday activities. CharadesDF was recorded using mobile phones leading to low/moderate quality videos compared to the more professionally captured DF40. Humans outperform AI detectors on both datasets, with the gap widening in the case of CharadesDF where AI accuracy collapses to near chance (0.537) while humans maintain robust performance (0.784). Human and AI errors are complementary: humans miss high-quality deepfakes while AI detectors flag authentic videos as fake, and hybrid human-AI ensembles reduce high-confidence errors. These findings suggest that effective real-world deepfake detection, especially in non-professionally produced videos, requires human-AI collaboration rather than AI algorithms alone.
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。