Eye_Illness_Det

AUTOMATED ASSISTIVE SCREENING OF RETINAL PATHOLOGY THROUGH ADVANCED NEURAL NETWORK [Classes Detected: Diabetic Retinopathy, Glaucoma, Cataract & Normal.] Ask For The Weight Files at [email protected] .

Contents

• Evaluation Metrics
• Confusion Matrix
• Result and Graphs
• Validation Batch Images

Evaluation Metrics

Metrics	Precision (P)	Recall (R)	mAP50	mAP50-95
Overall	0.908	0.936	0.966	0.964
Cataract	0.927	1.000	0.921	0.978
Diabetic	0.977	0.995	0.971	0.995
Glaucoma	0.874	0.900	0.947	0.945
Normal	0.830	0.944	0.943	0.943

Confusion Matrix:

Figure 1.1: Confusion Matrix depicting the model's performance on selected classes.

Result and Graphs:

1. 

Figure 2.1: Results.

2. 

Figure 2.2: F1 Curve.

3. 

Figure 2.3: Precision-Recall Curve.

4. 

Figure 2.4: Precision Curve.

5. 

Figure 2.5: Recall Curve.

Validation Batch Images:

1. 

Figure 3.1: Ground Truth Labels for Validation Image 1.

2. 

Figure 3.2: Predicted Labels for Validation Image 1.

3. 

Figure 3.3: Ground Truth Labels for Validation Image 2.

4. 

Figure 3.4: Predicted Labels for Validation Image 2.

5. 

Figure 3.5: Ground Truth Labels for Validation Image 3.

6. 

Figure 3.6: Predicted Labels for Validation Image 3.

Thank you for using the project!