DeepFake Detection on FF++ Average across scenarios benchmark leaderboard · SOTA2 Research

DeepFake Detection on FF++ Average across scenarios

95.7AccuracyLR-GCN

Evaluation Results

Method	Links
LR-GCN2025.12		95.7	93.6	97.4
TALL-Swin2025.12		82.1	66.4	68.3
Xception2025.12		81.3	59.9	63.6
CORE2025.12		81.2	56.2	62.2
RECCE2025.12		81.1	56	62.3
F3-net2025.12		80.4	56.9	62.3
DFIL2025.12		79.7	72.1	71.8
UCF2025.12		79.3	59.9	64.2
MaskRelation2025.12		78.8	77.5	87.7
DFGaze2025.12		78.5	63.4	69.1