Table 1.
Comparison of existing literature.
| Paper | Model | Datasets | Number of Image | Image Size (Pixels) |
|---|---|---|---|---|
| Shanthi et al. [3] | Modified AlexNet | Messidor | 1190 | 259 × 259 |
| Hemanth et al. [10] | Proposed CNN model | MESSIDOR | 400 | 150× 255 |
| Gu et al. [11] | Proposed transformer model by using vision transformers and residual attention | DDR and IDRiD |
13,673 and 516 |
512 × 512 |
| Sehrish et al. [12] | Ensemble Classifier | DR’s Kaggle dataset | 35,126 | 512 × 512 |
| Liu et al. [13] | Hybrid deep learning model | DeepDR, APOTS, and EyePACS | 1200, 3662 and 35,126 |
380 × 380 |
| Wu et al. [14] | CF-DRNet | IDRiD and DR’s Kaggle dataset | 413 and 35,126 | 256 × 256 |
| Lam et al. [15] | CNN based architectures | DR’s Kaggle dataset and Messidor-1 | 35,126 and 1090 | 256 × 256 |
| Gao et al. [16] | Inception@4 | DR Fundus | 4476 | 600 × 600 |