Table 7 Evaluation results of the comparison against the most recent research endeavors in the field of Arabic character classification in artificial intelligence.
Study | Dataset | Methods | Accuracy |
|---|---|---|---|
Alghamdi, Taghreed, et al.19, 2021 | Arabic OCR dataset | CNN and VGG16 | 92.09% |
A. Mohammed and R. Kora16, 2022 | HODA Arabic handwritten digit dataset | Ensemble CNN and LSTM | 99.39% |
K. M. O. Nahar et al.21, 2023 | Arabic OCR dataset | VGG16, VGG19 | 88.8% |
M. S. Alwagdani, et al.15, 2023 | Hijja dataset and AHCR | SVM | 92.96% |
Alsayed. A, et al.18, 2025 | AHCR | CNN + Kolmogorov Arnold Networks | 97.71% |
M. G. Mahdi, et al.7, 2024 | AHCR dataset | Bi-lstm | 97.05% |
H. AlShehri24, 2024 | AHCR dataset | Transfer learning | 98.11% |
Our proposed work | Arabic OCR Dataset | A hybrid transformer ViT model | 99.51 |
AHCR dataset | 98.20 |
