Vision Transformer-Based Video Hashing Retrieval for Tracing the Source of Fake Videos

Table 3

Comparison of accuracy (ACC) with existing methods on multiple datasets.

Methods	Celeb-DF	DFD	FF++ (HQ)
Xception [54]	0.994	0.831	0.994
Face X-ray [52]	0.996	0.856	0.960
Grad-CAM [53]	0.794	0.919	0.992
STIL [55]	0.996	—	0.986
ISTVT [3]	0.998	—	0.996
MRL [56]	0.999	—	0.938
Ours	0.994	0.963	0.999

Bold values represent the best results in the correlation dataset.