Table 4 Comparative Performance Metrics of Baseline Models and GWO-CTransNet on ASL Datasets (Mean ± Std over 30 runs).

From: A hybrid CNN-transformer framework optimized by Grey Wolf Algorithm for accurate sign language recognition

Model

Accuracy %

Precision %

Recall %

F1-score %

MCC

AUC

Baseline CNN

92.13 ± 0.27

91.87 ± 0.32

91.45 ± 0.28

91.65 ± 0.30

0.890

0.943

Transformer Only

94.22 ± 0.19

93.75 ± 0.25

93.90 ± 0.24

93.82 ± 0.22

0.915

0.962

CTransNet(No GWO)

96.07 ± 0.14

95.88 ± 0.21

95.91 ± 0.18

95.89 ± 0.16

0.943

0.973

GWO- CTransNet

99.40 ± 0.11

99.36 ± 0.08

99.27 ± 0.09

99.31 ± 0.07

0.988

0.992

Back to article page