Multiclass classification of eye diseases based on fundus images using CNN
| No. | Content | Notebook |
|---|---|---|
| 1 | Initial data exploration and analysis. | 01_eye_dis_detect.ipynb |
| 2 | Data preparation | 02_data_preprocessing.ipynb |
| 3 | Multi-class classification on the whole dataset with simple augmentation | 03_classification_attempt.ipynb |
| 4 | Binary-class classification: Normal vs Cataract |
04_binary_classification.ipynb |
| 5 | Multi-class classification: Normal vs Cataract vs Myopia |
05_multiclass_classification.ipynb |
| 6 | ConvNet filters analysis | 06_filter_analysis.ipynb |
| 7 | Simple tool for image augmentation | augmentor.ipynb |
| Content | Notebook |
|---|---|
Final model (Normal vs Cataract vs Myopia) |
model5_RGB_n10.h5 |
| FINAL - whole project in one notebook (notebooks 1-6 included) | ocular_disease_FINAL.ipynb |
| Project presentation | project_presentation.pdf |