Table 3.

Detailed results of Task I. Precision, recall, F1 score and their mean values are reported for each team.

(a) VisionLabs
	Precision	Recall	F1-score
Real	0.89	0.88	0.89
Deepfake	0.95	0.96	0.96
accuracy			0.94
macro avg	0.92	0.92	0.92
weighted avg	0.94	0.94	0.94
(b) DC-GAN (Amped Team)
	Precision	Recall	F1-score
Real	0.86	0.78	0.82
Deepfake	0.92	0.95	0.93
accuracy			0.90
macro avg	0.89	0.87	0.88
weighted avg	0.90	0.90	0.90
(c) Team Nirma
	Precision	Recall	F1-score
Real	0.55	0.80	0.65
Deepfake	0.90	0.74	0.81
accuracy			0.75
macro avg	0.72	0.77	0.73
weighted avg	0.80	0.75	0.76
(d) AIMH Lab
	Precision	Recall	F1-score
Real	0.52	0.49	0.51
Deepfake	0.80	0.82	0.81
accuracy			0.73
macro avg	0.66	0.66	0.66
weighted avg	0.72	0.73	0.72
(e) PRA Lab—Div. Biometria
	Precision	Recall	F1-score
Real	0.43	0.76	0.55
Deepfake	0.86	0.59	0.70
accuracy			0.64
macro avg	0.64	0.68	0.62
weighted avg	0.74	0.64	0.66
(f) Team Wolfpack
	Precision	Recall	F1-score
Real	0.05	0.06	0.05
Deepfake	0.59	0.55	0.57
accuracy			0.41
macro avg	0.32	0.30	0.31
weighted avg	0.44	0.41	0.42
(g) SolveKaro
	Precision	Recall	F1-score
Real	0.17	0.31	0.22
Deepfake	0.59	0.39	0.47
accuracy			0.37
macro avg	0.38	0.35	0.34
weighted avg	0.47	0.37	0.40