rPPG 深度伪造检测器学到的并非心跳信号
Pulse or Provenance? What rPPG-Based Deepfake Detectors Learn
做伪造检测的研究者该看看这篇：它证明了不少高 AUC 的 rPPG 检测器学到的是视频来源指纹，不是心跳，连官方 Celeb-DF 数据集的训练测试划分都有泄漏。
arXiv 论文（编号 2611.13992）检验了 rPPG（远程光电容积脉搏波）类深度伪造检测器在 Celeb-DF++（22 种伪造方法）和 FaceForensics++ 上的学习机制。结果显示，简单的色度信号在说话人脸伪造上 AUC 达 0.998，超过专用 rPPG 提取器的 0.87，但该信号并非生理性来源。实验发现信号能量集中在 0.7–4 Hz 心率带以上，从背景区域提取同样有效，打乱帧序即可消除，而注入合成脉搏几乎没有作用。
作者定位了信号来源：在两种生成器的自重演实验中，从单帧驱动人脸动画会引入同一高频信号，而逐帧独立重建则不会。基于人脸的色度检测器在 Celeb-DF++ 训练后，将超过 99% 的 FF++ 真实视频误判为伪造，CLIP 类检测器同样如此，说明它们编码的主要是视频来源而非篡改痕迹；冻结的 CLIP 骨干网络单独表现就已匹敌或超过与 rPPG、频域特征的融合方案。
论文还发现官方 Celeb-DF 与 Celeb-DF++ 的训练/测试划分存在身份与源视频重叠，自定义协议复用测试视频和身份可使较弱检测器虚高最多 0.30 AUC。
Pulse or Provenance? What rPPG-Based Deepfake Detectors Learn
Remote photoplethysmography (rPPG) has been proposed as a physiological cue for deepfake detection, based on the premise that synthesized faces lack a genuine cardiac signal. We examine what rPPG-style detectors learn on Celeb-DF++ (22 forgery methods) and FaceForensics++ (FF++), using one identically sampled window per real and fake video. On talking-face forgeries, simple chrominance traces reach an AUC of 0.998, exceeding a dedicated rPPG extractor (0.87). Controlled experiments indicate that this signal is not physiological in origin. On fully synthesized forgeries, it is equally discriminative when extracted from background regions or from full frames without face detection, and it is concentrated above the 0.7 to 4 Hz heart-rate band. It is largely removed by shuffling frame order but persists after normalizing resolution, frame rate and encoding, and an independent reimplementation reproduces these findings. Removing high-frequency content makes most synthesized videos appear real, adding it makes real videos appear fake, and injecting a synthetic pulse has little effect. In controlled self-reenactment experiments with two generators, animating a face from a single frame introduces the same high-frequency signal, whereas reconstructing each frame from its own source does not. Face-based chrominance detectors trained on Celeb-DF++ classify over 99% of real FF++ videos as fake, indicating that they largely encode video source rather than manipulation, and a CLIP-based detector exhibits the same behavior. A frozen CLIP backbone matches or exceeds fusions of CLIP with rPPG and frequency features. Finally, we show that the official Celeb-DF and Celeb-DF++ splits share identities and source-video content between training and test, and that custom protocols reusing test videos and identities in training inflate weaker detectors by up to 0.30 AUC.
本条情报归属 AITOP「论文」核心追踪栏目。在跨区域产业观察坐标中，该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线，反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括：Deepfake检测、rPPG、Celeb-DF++、FaceForensics++，读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。
信源核验说明：本文原始线索采编自「arXiv cs.LG」（@Othmane Harraq, Tamer Aldwairi），经 AITOP 多维评分管线结构化提炼。