Table 2 Classifier outcome of ECMVFD-FTLTDO technique under dissimilar epochs on GRIP dataset.

From: A multi-model feature fusion based transfer learning with heuristic search for copy-move video forgery detection

Class

\(Acc{u}_{y}\)

\(Pre{c}_{n}\)

\(Rec{a}_{l}\)

\({F1}_{score}\)

\(MCC\)

Epoch—500

Tampered

90.09

99.34

90.09

94.49

89.46

Original

99.34

90.15

99.34

94.52

89.46

Average

94.72

94.74

94.72

94.51

89.46

Epoch—1000

Tampered

89.79

98.36

89.79

93.88

88.15

Original

98.36

89.79

98.36

93.88

88.15

Average

94.07

94.07

94.07

93.88

88.15

Epoch—1500

Tampered

92.49

98.09

92.49

95.21

90.43

Original

98.03

92.26

98.03

95.06

90.43

Average

95.26

95.17

95.26

95.13

90.43

Epoch—2000

Tampered

66.07

99.55

66.07

79.42

68.98

Original

99.67

72.84

99.67

84.17

68.98

Average

82.87

86.19

82.87

81.79

68.98

Epoch—2500

Tampered

91.59

90.24

91.59

90.91

80.80

Original

89.14

90.64

89.14

89.88

80.80

Average

90.37

90.44

90.37

90.40

80.80

Epoch—3000

Tampered

83.78

98.94

83.78

90.73

83.26

Original

99.01

84.79

99.01

91.35

83.26

Average

91.40

91.86

91.40

91.04

83.26