arXiv：cs.LG· Danil Davydov, Bader Rasheed, Dmitriy Vatolin·· 3 小时前
光照先验对换脸检测有帮助吗？Temporal Self-Blended Images 的对照研究
Does an Illumination Prior Help Face-Swap Detection? A Controlled Study of Temporal Self-Blended Images
AI 导读
Temporal Self-Blended Images（T-SBI）通过在同一视频帧间迁移光照统计量来训练换脸检测器，但研究发现并无光照特异性提升：四个数据集上的 AUC 差异均在种子波动范围内，506,328 个属性分桶样本也未显示强光下错误率优先下降。
正文
Abstract: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 ({\Delta}L). Using five training regimes and a three-seed comparison of high- and low-{\Delta}L 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.
| Subjects: | Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.11706 [cs.LG] |
| (or arXiv:2610.11706v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.11706 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Bader Rasheed [view email]
[v1]
Thu, 8 Oct 2026 11:12:45 UTC (387 KB)
来源：arXiv：cs.LG · arxiv.org