Table 3 Performance metrics for discrimination of twelve wood species using standard images

From: Inferring ancient from modern: a deep learning approach for species identification of archaeological wood

Species

Recall (%)

Precision (%)

Genus-level average recall (%)

Genus-level average precision (%)

Betula albosinensis

99.95

98.24

99.95

98.24

Catalpa ovata

100.00

100.00

100.00

100.00

Cinnamomum camphora

100.00

96.98

100.00

96.98

Kalopanax septemlobus

100.00

100.00

100.00

100.00

Phoebe bournei

88.90

76.64

88.25

89.21

Phoebe chekiangensis

99.15

99.23

Phoebe hungmaoensis

85.81

98.22

Phoebe nanmu

89.68

81.24

Phoebe neurantha

51.00

99.22

Phoebe sheareri

90.42

92.95

Ulmus pumila

100.00

99.30

100.00

99.30

Zelkova schneideriana

99.48

100.00

99.48

100.00

Overall Accuracy

94.20