发表机构
Innopolis University（因诺波利斯大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究通过对照实验发现，时间自混合图像（T-SBI）的光照先验未提升人脸换脸检测的光照特异性性能，但可改变预测阈值并增强对JPEG压缩的鲁棒性，凸显训练方法评估需结合目标与阈值效应。
AI 中文摘要
自混合图像被广泛用于训练人脸换脸检测器，但主要捕捉的是混合伪影。本研究探究加入光照不一致性是否能提升检测效果。时间自混合图像（Temporal Self-Blended Images，T-SBI）在同一视频的帧之间传递光照统计信息，其不匹配程度由亮度差（ΔL）控制。通过五种训练方案及高ΔL与低ΔL训练的三次随机种子对比，我们未发现光照特异性提升的证据：在四个数据集上，AUC差异处于随机种子的变异范围内；对506328个按属性分组的样本分析显示，在恶劣光照下错误未出现优先减少。相反，T-SBI会改变预测分数，使FaceForensics++上的最优阈值变化约0.34，Celeb-DF上变化约0.30，导致固定阈值下的比较具有误导性。不过，T-SBI提升了DFDC数据集对重度JPEG压缩的鲁棒性（AUC为0.780，对比0.696），这可能反映其更依赖低频线索。这些发现强调了针对预期目标评估训练方法并考虑阈值效应的重要性。
英文摘要
Self-blended images are widely used to train face-swap detectors, but primarily capture blending artifacts. We investigate whether adding illumination inconsistencies improves detection. Temporal Self-Blended Images (T-SBI) transfer lighting statistics between frames of the same video, with the mismatch controlled by luminance difference (ΔL). Using five training regimes and a three-seed comparison of high- and low-ΔL training, we find no evidence of illumination-specific improvements. AUC differences remain within seed variability across four datasets, and an analysis of 506,328 attribute-binned samples shows no preferential reduction in errors under harsh lighting. Instead, T-SBI shifts prediction scores, changing optimal thresholds by approximately 0.34 on FaceForensics++ and 0.30 on Celeb-DF, making comparisons at a fixed threshold misleading. However, T-SBI improves robustness to heavy JPEG compression on DFDC (AUC 0.780 versus 0.696), potentially reflecting greater reliance on low-frequency cues. These findings highlight the importance of evaluating training methods against their intended targets and accounting for threshold effects.
