|
| Classifier/Method | Interface | Dataset type/size | Accuracy % |
| Proposed Work | SVM XGB KNN ET | Sensor-based glove | Arabic-added language 30 different static and dynamic gestures Original dataset size 1,029,801 10 performers – 11 features | 90.17 91.5 98.68 99.54 |
| Others | [4] ANN RF KNN DT SVM | Sensor-based glove | Malaysian sign language 21 static gestures 4 performers – 64 features data size 2100 | 99 99 99 97 89 |
| [5] CNN CapsNet 3 rout. CapsNet 5 rout. | Wearable IMU system | Indian Sign Language 20 sentences with each consisting of 2 - 4 signs- 10 performers | 87.99 94 92.5 | |
| [8] Multi-label Classification LP-MC | Wearable sensors | Indian sign language 100 isolated signs 10 performers – 20,000 samples | No acc. Categ. error of 2.73% |