Skip to main content

Table 1 Gender confusion tables

From: A study on different experimental configurations for age, race, and gender estimation problems

 

(a) LBP—unbalanced

 

(b) LBP—balanced

 

M T (%)

F T (%)

 

M T (%)

F T (%)

\({M_{P}^{N}}\)

97.9

2.1

\({M_{P}^{N}}\)

93.8

6.2

\({F_{P}^{N}}\)

14.1

85.9

\({F_{P}^{N}}\)

5.2

94.8

\({M_{P}^{S}}\)

98

2

\({M_{P}^{S}}\)

73.5

26.5

\({F_{P}^{S}}\)

14.7

85.3

\({F_{P}^{S}}\)

28.4

71.6

TA N

91.9

TA N

94.3

TA S

91.6

TA S

72.5

(c) HOG—unbalanced

(d) HOG—balanced

 

M T (%)

F T (%)

 

M T (%)

F T (%)

\({M_{P}^{N}}\)

98.8

1.2

\({M_{P}^{N}}\)

92.4

7.6

\({F_{P}^{N}}\)

21.5

78.5

\({F_{P}^{N}}\)

4.7

95.3

\({M_{P}^{S}}\)

97.9

2.1

\({M_{P}^{S}}\)

76.4

23.6

\({F_{P}^{S}}\)

14.7

85.3

\({F_{P}^{S}}\)

25.6

74.4

TA N

88.6

TA N

93.8

TA S

91.6

TA S

75.4

(e) SWLD—unbalanced

(f) SWLD—balanced

 

M T (%)

F T (%)

 

M T (%)

F T (%)

\({M_{P}^{N}}\)

97.7

2.3

\({M_{P}^{N}}\)

92

8

\({F_{P}^{N}}\)

17.8

82.2

\({F_{P}^{N}}\)

7.5

92.5

\({M_{P}^{S}}\)

97.7

2.3

\({M_{P}^{S}}\)

83.2

16.8

\({F_{P}^{S}}\)

17.7

82.3

\({F_{P}^{S}}\)

20.4

79.6

TA N

90

TA N

92.3

TA S

90

TA S

81.4

(g) CLBP—unbalanced

(h) CLBP—balanced

 

M T (%)

F T (%)

 

M T (%)

F T (%)

\({M_{P}^{N}}\)

98.1

1.9

\({M_{P}^{N}}\)

94.3

5.7

\({F_{P}^{N}}\)

11.4

88.6

\({F_{P}^{N}}\)

5.4

94.6

\({M_{P}^{S}}\)

98.2

1.8

\({M_{P}^{S}}\)

88

12

\({F_{P}^{S}}\)

11.8

88.2

\({F_{P}^{S}}\)

12.2

87.8

TA N

93.4

TA N

94.5

TA S

93.2

TA S

87.9

  1. Each table presents the results for a specific descriptor for both scaled and non-scaled approaches. Moreover, the tables in the left column have been trained on unbalanced dataset; the tables in the right column over the balanced one. M and F are for male and female; subfix P and T indicate predicted and true value, respectively. TA is for total accuracy; superfix N and S refer to the non-scaled and scaled approach, respectively