Local Corner and Motion Key Point Trajectory Extraction for Facial Forgery Identification

Table 3

AUC (%) results of the proposed method and mainstream methods on the FF++ dataset.

Methods	FF++
	Raw	c40
Xception [18]	99.7	86.5
Face X-ray [23]	99.1	61.6
LRNet [11]	99.9	95.7
DSP-FWA [24]	93.0	—
Capsule [25]	96.6	—
Single XceptionNet [26]	—	97.8
Chen et al. [27]	99.92	95.2
SPSL [28]	—	82.8
PCL + I2G [29]	99.79	—
FTCN [30]	99.7	—
Lip forensics [31]	98.9	94.2
FDFL [32]	99.7	92.4
Ours	99.9	97.6

The bold values are used to highlight the experimental records that represent the most optimal performance within each experimental group. Specifically, the bold values labeled as “Ours” indicate the results obtained from the experiments conducted using our proposed method.