Table 6 Comparative performance analysis of GWO-CTransNet against existing sign language recognition methods on ASL alphabet and ASL MNIST datasets.
References | Approach used | Performance metric | ASL alphabet | ASL MNIST |
|---|---|---|---|---|
AEGWO-NET-ANN | Accuracy | 97.69 | 98.26 | |
F1-score | 95.06 | 96.83 | ||
MCC | 93.82 | 97.46 | ||
AUC | 96.07 | 97.83 | ||
HOG-GA-MLP | Accuracy | 91.34 | 94.35 | |
F1-score | 89.72 | 94.32 | ||
MCC | 90.16 | 93.39 | ||
AUC | 89.34 | 92.67 | ||
LBP-GA-SVM | Accuracy | 91.12 | 96.71 | |
F1-score | 89.96 | 94.08 | ||
MCC | 90.16 | 96.59 | ||
AUC | 89.64 | 97.02 | ||
GWO-rough set-Naïve bayes | Accuracy | 89.06 | 97.35 | |
F1-score | 87.43 | 95.12 | ||
MCC | 88.62 | 95.52 | ||
AUC | 87.59 | 96.05 | ||
PSO-CNN | Accuracy | 96.45 | 99.16 | |
F1-score | 94.43 | 98.45 | ||
MCC | 92.99 | 99.57 | ||
AUC | 95.91 | 99.73 | ||
GWO-CNN | Accuracy | 98.42 | 99.83 | |
F1-score | 96.87 | 97.36 | ||
MCC | 95.04 | 98.42 | ||
AUC | 97.28 | 99.06 | ||
Proposed Approach | GWO-CTransNet | Accuracy | 99.40 | 98.07 |
F1-score | 99.31 | 97.90 | ||
MCC | 98.80 | 97.50 | ||
AUC | 99.20 | 97.87 |
