Table 3. Comparison with existing work.

Reference	Proposed algorithm	Accuracy
Amin et al. (2018)	Fusedbased ensemble method	86%
Mallick et al. (2019)	Deep Neural Network	89%
Badža & Barjaktarović (2020)	CNN	95.40%
Hashemzehi et al. (2020)	CNN and NADE	95%
Bodapati et al. (2021)	Deep-CNN	95.23%
Nallamolu et al. (2022)	ViT Model	93.48%
Isunuri & Kakarla (2023)	Ensemble CNN	96.95%
Khan et al. (2023)	Ensemble XG-Ada-RF model	95.9%
Şahin, Özdemir & Temurtaş (2024)	BMO-ViT	96.6%
Proposed	Swin Transformer	97%